Motors Involved in Transport

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By

Lina Wang

Graduate Program in

Neuroscience Graduate Studies Program

The Ohio State University

2011

Dissertation Committee:

Anthony Brown, Advisor

Jeff Kuret

Christine Beattie

Chen Gu

Copyright by

Lina Wang

2011

Abstract

The first paragraph was reproduced from the Wang and Brown 2010 paper, which was written by both authors.

Neurofilaments are synthesized in the nerve cell body and transported bidirectionally along axons by slow axonal transport. Direct observations of neurofilament transport indicate that the slow rate is due to rapid movements interrupted by prolonged pauses.

Previous studies suggest that the neurofilament motors include / and -1A (KIF5A). Recently, mutations in kinesin-1A have been linked to one of the dominantly inherited forms of hereditary spastic paraplegias (SPG10), suggesting that neurofilament transport may be disrupted in SPG10. To explore this hypothesis, I investigated the effect of an SPG10 point mutant, N256S-kinesin-1A, on neurofilament transport. Previous studies have shown that this mutant has impaired velocity and processivity in vitro. I transfected cultured mouse cortical neurons with GFP-tagged neurofilament M (GFP-NFM), with or without mutant or wild type kinesin-1A, and then analyzed neurofilament transport in axons using live-cell time-lapse imaging after 8-12 days in culture. The N256S mutant decreased anterograde neurofilament transport flux (from 136 to 35 µm/axon/hr) by decreasing anterograde transport frequency (from 3.9 to 1.0 moving filaments/hr) without any statistically significant

ii effect on anterograde transport velocity. Consistent with previous observations from the

Brown lab on neurofilament transport in kinesin-1A knockout neurons, retrograde neurofilament transport flux was also decreased (from 111 to 85 µm/axon/hr). This was due to decreased retrograde transport frequency (from 3.1 to 1.9 moving filaments/hr) in spite of a statistically significant increase in retrograde transport velocity (from 0.32 to

0.41 µm/s). I conclude that the N256S-kinesin-1A mutant disrupts bidirectional movement of neurofilament in cultured mouse cortical neurons.

I also studied the interaction between and motors. While there is evidence for an interaction between neurofilaments and dynein/dynactin, there is no such evidence for kinesin-1A. Therefore, I examined the interaction of kinesin-1A and dynein/dynactin

(p150 subunit) with neurofilaments using immunoprecipitation with magnetic beads.

About 0.1% of the total kinesin-1A and about 0.5% of the total p150 co- immunoprecipitated with neurofilaments from mouse brain homogenate. In addition, about 0.1% of the neurofilament protein co-immunoprecipitated with kinesin-1A, and about 1.0% with p150. To test the specificity of the interaction, I designed a sequential immunoprecipitation approach, in which I first precipitated all the neurofilaments, and then I performed a second immunoprecipitation with the same antibody using the immuno-depleted supernatant from the first immunoprecipitation. I found that neurofilament antibody failed to precipitate kinesin-1A and p150 from the neurofilament- depleted supernatant in the second immunoprecipitation, even though both kinesin-1A and p150 were present in abundance. The reciprocal sequential immunoprecipitation

iii experiment, using kinesin-1A or p150 antibody, yielded similar results. I also found that kinesin-1A and p150 also co-purified and co-immunoprecipitated with neurofilaments from mouse and rat spinal cords. Finally, I found that kinesin-1A interacted with each neurofilament subunit in HEK cells. Together, these data suggest a specific interaction between neurofilaments and kinesin-1A and confirm the previously reported interaction between neurofilaments and dynein/dynactin.

In conclusion, I present several lines of evidence suggesting that kinesin-1A is a neurofilament motor. Mutations in kinesin-1A, such as SPG10 mutations, may disrupt neurofilament transport in patients, which may contribute to the etiology of this disease.

iv

Dedicated to my family…

v

Acknowledgments

First and foremost, I would like to thank my advisor Dr. Anthony Brown for his patience, support and invaluable advice. It would be impossible for me to complete my doctoral study without his support, encouragement and guidance. He, as an example, teaches me how to be a great scientist. He leads my way to become a professional scholar, and for that I am grateful.

I would also like to thank my committee members, Dr. Jeff Kuret, Dr. Christine Beattie and Dr. Chen Gu for their patience, insightful suggestions and critical reviews of my dissertation.

I would like to thank past and present members of the Brown lab who have made my graduate study enjoyable. My thanks go out to Niraj Trivedi, Atsuko Uchida, Nael Alami,

Gulsen Colakoglu, Paula Monsma and Cynthia Neuendorf.

I have been blessed to have a group of beloved friends during my stay in Columbus, whose support and love I will always appreciate. My thanks go out to all of you,

vi especially to Mingxi Yin, Sui Huang, Haiyan Peng, Ruifeng Cao and Yuan Lin. In addition, I would like to thank Junying Yu and Yajing Zhao. I am so blessed to have all of you in my life.

Last but not least, I would like to thank my father and mother, who brought me to the world. I am grateful for their unconditional love. They are always there to support me when I am weak. They teach me how to be a good person. For all of these, I am grateful.

vii

Vita

January 06, 1983 ...... Born, Tianjin, China

2005...... B.S. Biological Sciences, NanKai University

2006 to present ...... Graduate Research Associate, Neuroscience

Graduate Studies Program, The Ohio State

University

Publications

Yuan L, Zheng Y, Zhu J, Wang L and Brown A. (2011) Object Tracking with Particle

Filtering in Fluorescence Microscopy Images: Application to the Motion of

Neurofilaments in Axons. IEEE Trans Med Imaging, Aug 22:99.

Wang L and Brown A (2010) A hereditary spastic paraplegia mutation in kinesin-

1A/KIF5A disrupts neurofilament transport. Molecular Neurodegeneration, Nov 18;5:52.

viii

Fields of Study

Major Field: Neuroscience

ix

Table of Contents

Abstract ...... ii

Dedicated to my family… ...... v

Acknowledgments...... vi

Vita ...... viii

LIST OF FIGURES ...... xiv

Chapter 1 Introduction ...... 1

1.1 Neurofilaments ...... 1

1.1.1 General introduction ...... 1

1.1.2 Neurofilament expression and assembly ...... 5

1.1.3 Function of neurofilaments ...... 9

1.2 Axonal Transport...... 14

1.2.1 General introduction ...... 14

1.2.2 Fast axonal transport ...... 18

1.2.3 Slow axonal transport ...... 19

1.2.4 Neurofilament phosphorylation and transport ...... 25

1.3 Molecular motors ...... 27

1.3.1 Kinesin superfamily (KIF)...... 28 x

1.3.2 Dynein/dynactin complex ...... 33

1.3.3 superfamily ...... 38

1.4 Neurofilaments and neurodegenerative diseases ...... 43

1.4.1 Amyotrophic Lateral Sclerosis ...... 44

1.4.2 Charcot-Marie-Tooth ...... 47

1.4.3 ...... 48

1.4.4 Diabetic neuropathy ...... 50

1.5 Axonal transport and neurodegenerative diseases ...... 52

Chapter 2 Materials and Methods ...... 56

2.1 Cell culture ...... 56

2.1.1 Astroglial cells and cortical neurons co-culture ...... 56

2.1.2 HEK 293T cell culture ...... 59

2.2 Cloning ...... 59

2.3 Transfection ...... 64

2.3.1 Electroporation ...... 64

2.3.2 Lipofection...... 64

2.4 Live cell imaging and analysis ...... 65

2.4.1 Imaging of neurofilament movement through gaps ...... 65

2.4.2 Analysis of neurofilament movement ...... 66

xi

2.4.3 Imaging of EB1-YFP movement ...... 68

2.4.4 Analysis of EB1-YFP movement ...... 68

2.5 Immunocytochemistry ...... 69

2.6 Immunoprecipitation (IP) ...... 71

2.6.1 Immunoprecipitation from mouse brain homogenate ...... 71

2.6.2 Immunoprecipitation from HEK293T cells ...... 73

2.7 Western Blotting ...... 74

2.8 Immunostaining of native neurofilaments and motor in vitro ...... 77

Chapter 3 The role of a hereditary spastic paraplegia mutation in kinesin-1A/KIF5A in neurofilament transport ...... 80

3.1 Introduction ...... 80

3.2 Observation of neurofilament transport ...... 85

3.3 Neurofilament transport frequency and flux ...... 88

3.4 Neurofilament transport bout velocity, distance and duration ...... 91

3.5 Comparison of neurofilament content of distal axons ...... 96

3.6 polarity in axons of cultured mouse cortical neurons ...... 99

3.7 Summary ...... 102

Chapter 4 Interaction of neurofilaments with kinesin-1A and dynein/dynactin ...... 105

4.1 Introduction ...... 105

xii

4.2 Characterization of kinesin-1 antibodies ...... 109

4.3 Interaction between neurofilaments and motor proteins revealed by conventional

immunoprecipitation (IP) from mouse brain homogenate ...... 112

4.4 Interaction between neurofilament and motor proteins revealed by sequential IP

from mouse brain homogenate ...... 119

4.5 Interaction between neurofilament and motors confirmed by IP using

neurofilaments purified from mouse/rat spinal cords ...... 122

4.6 Interaction between kinesin-1A and each neurofilament subunit ...... 126

4.7 Summary ...... 130

Chapter 5 Discussion ...... 133

5.1 Kinesin-1A and neurofilament transport ...... 133

5.1.1 N256S-kinesin-1A mutation and neurofilament movement ...... 133

5.1.2 Models of bi-directional cargo movement ...... 135

5.1.3 Interaction between kinesin-1A and neurofilaments ...... 139

5.2 Kinesin-1A and hereditary spastic paraplegia ...... 145

5.2.1 Neurofilament and HSP ...... 145

5.2.2 Other cargoes for kinesin-1A ...... 149

5.3 Future directions ...... 150

References ...... 156

xiii

LIST OF FIGURES

Figure 1.1 Microscopic and schematic pictures of cytoskeletal components ...... 2

Figure 1.2 Representation of neurofilament proteins ...... 5

Figure 1.3 assembly ...... 8

Figure 1.4 Schematic representations of pulse-chase experiments ...... 17

Figure 1.5 Structures of the major ...... 30

Figure 1.6 Schematic structure of dynein/dynactin ...... 36

Figure 1.7 Structures of three major myosin proteins...... 39

Figure 3.1 Sequence comparison of mouse and human kinesin-1A ...... 85

Figure 3.2 Neurofilament moving through a gap ...... 87

Figure 3.3 Effect of N256S-kinesin-1A on the frequency and flux of neurofilament movement ...... 90

Figure 3.4 Effect of N256S-kinesin-1A on the velocity, distance and duration of bouts of neurofilament movement ...... 93

Figure 3.5 Summary of kinetic data ...... 95

Figure 3.6 N256S-kinesin-1A does not deplete neurofilaments from distal axons ...... 98

Figure 3.7 in cortical neuron axons are plus-end distal ...... 101

Figure 4.1 Characterization of H2 antibody by Western blot...... 110

Figure 4.2 Characterization of H2 antibody in immunoprecipitation ...... 112 xiv

Figure 4.3 IP with NFM antibody or IgG coated beads ...... 115

Figure 4.4 Quantification of proteins in the pellet ...... 116

Figure 4.5 IP with H2 antibody or IgG coated beads ...... 117

Figure 4.6 IP with p150 antibody or IgG coated beads ...... 118

Figure 4.7 Sequential IP with NFM antibody ...... 120

Figure 4.8 Sequential IP with p150 or H2 antibody ...... 122

Figure 4.9 Comparison of 0.75 M NaCl treated and untreated groups ...... 125

Figure 4.10 Immunostaining of neurofilaments in NaCl treated and untreated groups.. 126

Figure 4.11 Interactions between kinesin-1A and neurofilament subunits ...... 129

Figure 5.1 Models for bidirectional cargo transport ...... 137

Figure 5.2 Immunostaining of purified neurofilaments with the neurofilament and kinesin-1A antibodies ...... 152

xv

Chapter 1 Introduction

1.1 Neurofilaments

1.1.1 General introduction

Neurofilaments are the most abundant cytoskeletal components in mature neurons. They play an important role in determining axon caliber, especially for large myelinated axons, which in turn affects the propagation of action potentials. Neurofilaments also function as mechanical shock absorbers to protect the cell from deformation. Neurofilaments belong to intermediate filament family, which is one of the three major cytoskeletal components in eukaryotic cells. The other two are ( filaments) and microtubules.

The diameter of intermediate filaments is 8-10 nm, which is intermediate between microfilaments (6 nm) and microtubules (24 nm) (Ishikawa et al., 1968) (Figure 1.1).

1

Figure 1.1 Microscopic and schematic pictures of cytoskeletal components consists of actin filaments, microtubules and intermediate filaments, which are different in their general organizations. (Bruce Alberts, 2002)

Intermediate filament family is classified into six types based on similarities in sequence and structure (Herrmann et al., 2007; Szeverenyi et al., 2008). Type I and type II intermediate filaments contain acidic and basic . They are expressed in epithelial cells and provide mechanical strength to cells. Type III intermediate filaments contain , , and glial fibrillary acidic proten. Vimentin is mainly expressed in mesenchymal cells. Desmin is present in muscle cells. Peripherin is expressed in peripheral neurons and glial fibrillary acidic proten is expressed in glial cells.

2

Type IV intermediate filaments include α-, , , and the three neurofilament subunits: NFL (low molecular weight, ~62 kDa), NFM (medium molecular weight, ~102 kDa) and NFH (high molecular weight, ~112 kDa). Due t the high content of negatively charged glutamic acids in their sequences and the extensive post-translational modifications, neurofilament subunits exhibit higher apparent molecular weights on SDS-PAGE (68 kDa for NFL, 160 kDa for NFM and 205 kDa for

NFH) (Lee et al., 1993). Neurofilaments were originally assumed to be formed by only

NFL, NFM and NFH subunits. However, recent studies revealed that α-internexin and peripherin were also co-assembled with neurofilament subunits (Beaulieu et al., 1999;

Yuan et al., 2006). In central nervous system, α-internexin is a fourth subunit of neurofilaments and peripherin is a fourth subunit of neurofilaments in peripheral nervous system.

Type V intermediate filaments include A/C, lamins B1 and B2. They are expressed in the nucleus. Type VI intermediate filaments contain Phakinin and Filensin, which are expressed in lens. They are proteins which cannot be categorized into one of

3 the other major intermediate filament classes. They are recognized as intermediate filaments based on their sequences and structures.

Like other intermediate filaments, neurofilament subunits have a tripartite structure, with a central α-helical rod domain flanked by non-helical amino head domain and carboxy terminal tail domain (Geisler et al., 1983) (Figure 1.2). This rod domain is highly conserved whereas the head and tail domains are less conserved. The head domain is enriched for serines and threonines. Phosphorylation of these sites plays an important role in regulating neurofilament assembly. The tail domain of the neurofilament is very distinctive. For NFL, the tail domain contains many glutamate residues which are called

E segments. The tail domains of NFM and NFH are much longer and contain numerous repeats of Lys–Ser–Pro (KSP) phosphorylation sites besides E segments. The Ser residues of KSPs are heavily phosphorylated in vivo (Julien and Mushynski, 1982; Julien and Mushynski, 1983), which has been suggested to play a role in neurofilament transport and regulation of axon calibers (Ackerley et al., 2003).

4

Figure 1.2 Representation of neurofilament proteins (A) Neurofilament subunit proteins with highly conserved rod domains flanked by head and tail domains. (B) Rotary shadowed view of a 10 nm native neurofilament. The core of the filament is formed by the head and rod domains of neurofilament subunits. The tail domains of NFM and NFH make up the “side-arms” projecting from the backbone. (Barry et al., 2007)

1.1.2 Neurofilament expression and assembly

Expression of the three neurofilament subunits is regulated at both transcriptional and post-transcriptional levels during neuronal development (Nixon and Shea, 1992;

Thyagarajan et al., 2007). During early development, NFL is expressed first followed by the expression of NFM (Carden et al., 1987; Willard and Simon, 1983). At this time, microtubule is the major cytoskeletal component in neurons (Cuenca et al., 1987). The expression level of NFL and NFM is relatively low. NFH starts to express after synaptogenesis, accompanied by rapidly increased NFL and NFM expression and

5 decreased microtubule expression (Pachter and Liem, 1984; Shaw and Weber, 1982).

Therefore, expression of the cytoskeletal proteins (neurofilaments and microtubules) seems to be finely balanced during the development of nervous system.

After expression, neurofilament subunits need to be assembled to form neurofilament polymers. The assembly of neurofilament depends on ionic strength, pH and temperature instead of nucleotide hydrolysis (Angelides et al., 1989). The first step of assembly is dimerization of NFL with either NFM or NFH to form parallel coiled coil dimmers through their rod domains. The second step is the formation of an anti-parallel tetramer from two dimers. Finally, the tetramers assemble to form protofilaments, which come together to form the 10 nm neurofilaments (Heins et al., 1993) (Figure 1.3). The tail domains of NFM and NFH form lateral projections from the filament backbone

(Hirokawa et al., 1984; Hisanaga and Hirokawa, 1988), which inter-connect neurofilaments with each other and cross-link neurofilaments to other cytoskeletal components, such as microtubules, and , such as mitochondria, to stabilized neurofilament networks.

6

NUDEL, a mammalian homologue of the Aspergillus nidulans nuclear distribution molecule NudE, has been shown to be involved in regulating neurofilament assembly. It interacted directly with NFL through the rod domain of NFL and indirectly with NFH, but did not interact with NFM. NUDEL promotes neurofilament assembly without co- assembling in mature neurofilament polymers. Knocking down of NUDEL has been shown to decrease NFL protein level and disrupt neurofilament assembly (Nguyen et al.,

2004). It is possible that the effect of NUDEL knocking down on neurofilament assembly might be caused by a reduction of NFL proteins instead of a direct effect of the reduction of NUDEL. Since then, there is no follow-up study on NUDEL and the neurofilament, therefore it is unclear whether NUDEL is truly involved in regulating neurofilament assembly or not.

7

Figure 1.3 Intermediate filament assembly Two monomers (A) are aligned in parallel to form a dimer (B) through their rod domains. Two dimers form an anti-parallel tetramer (C). Tetramers can associate with each other (D). Tetramers are packed together in a helical array in the final 10 nm intermediate filament (E). Upper left picture shows an electron micrograph of a final filament (Bruce Alberts, 2002).

8

In addition, neurofilament assembly can be regulated by post-translational modifications

(Ching and Liem, 1993; Ching and Liem, 1999; Gill et al., 1990; Wong and Cleveland,

1990). Phosphorylation of NFL head domain prevents neurofilament assembly and promotes disassembly of pre-existing neurofilaments (Hisanaga et al., 1990). It has been shown that protein kinase A and protein kinase C can phosphorylate NFL (Ser 51 and Ser

55), and NFM (Ser 23) to regulate neurofilament assembly (Sihag et al., 1999; Sihag and

Nixon, 1991). Besides phosphorylation sites, there are also O-linked N-acetyl glucosamine (O-GlcNAc) sites in NFM and NFH head domains (Dong et al., 1993; Dong et al., 1996), which locate close to the phosphorylation sites. It is possible that these post- translational modifications could influence each other to regulate neurofilament assembly.

1.1.3 Function of neurofilaments

Axonal caliber and conduction velocity

Neurofilaments as the prominent cytoskeletal component of large myelinated axons play a critical role in determining axon diameter, which in turn affects the conduction velocity of axons. The essential role neurofilaments played in axon radial growth has been well 9 documented. The first evidence demonstrating the role of neurofilaments in axon radial growth came from investigation of a mutant strain of the Japanese quail, called quiverer

(Quv). Quiverer has a nonsense mutation in NFL gene resulting in lacking of neurofilaments in axons because NFL is required for neurofilament polymerization. Both axon caliber and conduction velocity were significantly reduced in quiverer (Ohara et al.,

1993; Yamasaki et al., 1992; Yamasaki et al., 1991). In agreement with this finding, NFL knockout mice also developed axon atrophy due to lack of neurofilaments in axons (Zhu et al., 1997).

Studies from Eyer and Peterson also proved a critical role neurofilaments played in axon radial growth. In these studies, the authors derived transgenic mice expressing a NFH- beta-galactosidase fusion protein, in which the COOH terminus of NFH was replaced by beta-galactosidase. The fusion protein aggregated in perikarya of neurons resulting in depletion of neurofilaments from axons. Axons developed only 50% of normal diameter and conduction velocity was progressively reduced (Eyer and Peterson, 1994). Similar to these findings, the calibers of large axons in mutant mice in which one allele for each

10 neurofilament gene was disrupted were decreased from 5-9 μm to 1-5 μm. Neurofilament protein contents in these mice were reduced by 40% (Nguyen et al., 2000).

Since then, several studies have been conducted to determine the role of each neurofilament subunit in axon radial growth. As mentioned above, NFL is critical in determining axon calibers probably due to the fact that neurofilaments cannot polymerize without NFL. Deficiency in NFL represents a decrease in neurofilament polymers.

Besides NFL, NFM also plays an important role in determining axon caliber. Axons of mice with a null mutation in NFM had a dramatic reduction of NFL level and a slight increase of NFH level as well as increased numbers of microtubules. Axon calibers of both large and small diameter axons were decreased accompanied by a reduction of conduction velocity (Elder et al., 1998a). Moreover, the tail domain of NFM is essential for the radial growth of axons revealed by NFM tail-deleted (NF-MtailΔ) mutant mice

(Rao et al., 2003). In these mice, inhibited radial growth of axons was in company with reduced distance between neurofilaments without any change in neurofilament protein

11 content, suggesting disruptions of cross-bridges between neurofilaments since NFM tail domain is involved in side-arm formation. Although NFM tail domain is important in axon radial growth, phosphorylation of KSP repeats in NFM tail domain is not required for myelin-dependent radial growth of axons (Garcia et al., 2009). This was revealed by mutant mice in which all serines in KSP repeats of NFM had been replaced with alanines, which are phosphorylation incompetent. Axon caliber and motor neuron conduction velocity of these mice were unaltered compared with wild-type mice.

Several studies of NFH-/- mice revealed that NFH does not play a critical role in axonal radial growth (Elder et al., 1998b; Rao et al., 2002b; Rao et al., 1998; Zhu et al., 1998). In

NFH-/- mice, axons achieved comparable diameters as in wild-type mice in Rao and

Zhu’s studies. One possible explanation is that the effect of NFH deletion may be compromised by increased microtubule density and NFM levels, which were observed in these studies. In the study from Elder et al., the authors found that relatively large diameter axons were shifted to mid-sized axons and mid-sized axons were shifted to small axons in NFH-/- mice. One possible explanation for the divergences of this study from Rao and Zhu’s studies is that in Elder’s study, the authors did not observed any

12 changes in NFM levels or microtubule density, therefore loss of NFH may not be compensated by increased levels of other cytoskeletal components. Another possible is that in Elder’s study, the animals were examined at relatively younger age than in Rao’s study. In fact, Rao et al. found that deletion of NFH retarded the initial rate of radial growth of motor axons. However, wild-type axon diameters were achieved by 6 month.

In Elder’s case, the animals were examined at 4 month age and the axons were shifted to smaller diameters. It is possible that by 6 month, these axons may achieve wild-type axon diameters.

Dendritic Arborization

The role of neurofilaments in determining axon radial growth was extensively studied while their role in the development of dendrites was less investigated. Studies from Kong et al. showed that the ratio of NFM and NFH to NFL was critical for dendritic arborization in spinal motor neurons (Kong et al., 1998). In addition, NFL-/- mice exhibited inhibited dendritic growth most dramatically in large motor neurons (Zhang et al., 2002).

13

Mechanical role

Neurofilaments can form viscoelastic networks through cross-bridging between

NFM/NFH side-arms and other cytoskeletal components, such as microtubules. These viscoelastic networks are resistant to deformation (Leterrier et al., 1996). In vitro studies using atomic force microscopy (AFM) showed that neurofilaments could be stretched more than three-fold of their original lengths without breaking, suggesting that neurofilaments may indeed function as mechanical shock absorbers in vivo (Kreplak et al., 2005).

1.2 Axonal Transport

1.2.1 General introduction

A neuron is a highly polarized cell usually has a long slender axon and several dendrites.

Whereas dendrites are usually close to the cell body and have some capacity for protein synthesis, the axon is generally devoid of protein synthesis machineries. Therefore, materials destined to the axon are synthesized in the cell body and then exported to the axon. In contrast, materials destined to the cell body are transported from the end of the 14 axon to the cell body. This bi-directional movement is termed axonal transport, which continues throughout the life of the cell and is essential for axon function and survival.

The first evidence of axonal transport came from the classic ligation/accumulation study conducted by Weiss and Hiscoe in 1948 (Weiss and Hiscoe, 1948). In this study, the authors severed peripheral nerves in white rats and allowed them to regenerate through segments of transplanted arteries with lumens smaller than the diameters of the nerves.

The artery constricted the nerve passing through it. Overtime the authors observed axonal swellings proximal to the constrictions due to the accumulation of materials transported down the nerves. When the constrictions were removed, the swellings moved down the nerves at the rate of approximately 1 mm/day.

Most of our knowledge about the kinetics of axonal transport comes from radioisotopic pulse-chase studies, which allow the movement of proteins to be investigated (Grafstein and Forman, 1980) (Figure 1.4). In these studies, radio-labeled amino acids were injected into the vicinity of cell bodies, for example into dorsal root ganglion, which contains cell bodies of sciatic nerve. When these amino acids were taken up by the cell bodies, they

15 were incorporated into newly synthesized proteins and moved along axons. By injecting many animals at the same time and sacrificing them at different time points, the kinetics of transport can be analyzed. Using this method, two major classes of axonal transport were identified, fast axonal transport of membranous organelles and slow axonal transport of cytoskeletal and cytosolic proteins. The velocity of fast axonal transport ranges between 100-400 mm/day, while the velocity of slow axonal transport ranges between 0.2-5 mm/day (Grafstein and Forman, 1980; Lasek et al., 1984).

16

Figure 1.4 Schematic representations of pulse-chase experiments In pulse-chase experiments, radio-labeled amino acids are usually injected into dorsal root ganglion, where the cell bodies of sciatic nerve are located. Animals were killed at different time points (time 1, time 2, time 3 and time 4) after injections. Sciatic nerve was dissected out and cut into 5 mm segments for gel electrophoresis analysis. At time 1, the majority of labeled proteins were close to the cell body in segment 1, only a proportion of fast moving proteins (red and blue dots) appeared in segment 2. More and more labeled proteins were detected distally to the injection site with increased chase time. The red, blue and purple dots represent different groups of proteins that were transported down the axon at different rates, red most rapidly, purple most slowly (Harvey Lodish, 2000).

For a very long time, it has been believed that fast and slow axonal transport have fundamentally distinct mechanisms, but recent live-cell imaging studies suggest that slow axonal transport could be a result of fast intermittent movement (Roy et al., 2000; Wang

17 et al., 2000). Therefore, the difference between fast and slow axonal transport may not be the rate of transport, but rather the manner of the movement.

1.2.2 Fast axonal transport

The principal cargoes of fast axonal transport are membranous organelles. Proteins, lipids and polysaccharides move along the axon at fast rate by association with organelles or vesicles. Golgi-derived vesicles are transported from the cell body to the end of the axon

(anterogradely) at rates of 200-400 mm/day (2-5 μm/s). Endocytic vesicles, lysosomes and autophagosomes are transported from the end of the axon to the cell body

(retrogradely) at rates of 100-250 mm/day (1-3μm/s) (Grafstein and Forman, 1980). The general principles underlying fast axonal transport are relatively clear now. Membranous organelles are moving along microtubules and microfilaments. The movement is driven by motor proteins (see section 1.3) (Goldstein and Yang, 2000). The long distance movement along the axis of the axon uses microtubules as tracks. Microtubule plus end- directed kinesin motors drive anterograde transport, while minus end-directed kinesin and dynein motors drive retrograde transport. Transport along

18 tracks is powered by myosin motors. Membranous organelles interact either directly with the motors or indirectly with the motors through adaptor proteins (Karcher et al., 2002).

Direct observations of membranous organelle movement in live cells reveal that they move highly efficiently in a continuous and unidirectional manner at instantaneous rates that are comparable to the maximal rate of fast axonal transport determined by radioisotopic pulse-chase study. One exception to this notion is mitochondrial transport

(Hollenbeck, 1996), which are transported at rates of 20-70 mm/day (0.2-0.8 μm/s)

(Grafstein and Forman, 1980; Lorenz and Willard, 1978). The slower rate is likely due to their intermittent, bi-directional moving behavior (Ligon and Steward, 2000; Morris and

Hollenbeck, 1993), which is less efficient than the continuous and unidirectional movement of membranous organelles.

1.2.3 Slow axonal transport

Radioisotopic pulse-chase studies revealed two kinetically distinct slow axonal transport components (Lasek et al., 1984). Slow component a (SCa) contains neurofilaments, microtubules and a number of cytosolic proteins that move at rates of 0.2-1 mm/day 19

(0.002-0.01 μm/s). Slow component b (SCb) contains microfilaments and several hundred other cytosolic proteins that move at rates of 2-8 mm/day (0.02-0.09 μm/s), which is slightly faster than SCa. The mechanism by which these cytosolic proteins move is not well understood. One hypothesis is that cytoskeletal polymers may be carriers for slow axonal transport of cytosolic proteins (Lasek, 1986; Lasek et al., 1984). In support of this idea, several cytosolic proteins have been observed to associate with cytoskeletal polymers and transport at the same rate as cytoskeletal polymers (Knull and Walsh, 1992;

Macioce et al., 1999). However, the mechanism by which cytoskeletal proteins move has been controversy for nearly three decades.

Numerous attempts to observe slow axonal transport of cytoskeletal proteins in live cells were based on the assumption that slow axonal transport was a slow and synchronous movement. Many of these studies were based on fluorescence photobleaching or photoactivation strategies in which fluorescent labeled cytoskeletal proteins were injected into neurons and then a population of these proteins were marked by bleaching or activating the fluorescence in a narrow region along the axon with expectation to observe a slow translocation of the marked region. However, the marked regions in these studies

20 appeared to be stationary. In these studies, people observed gradual recovery of the bleached zones (Lim et al., 1990; Lim et al., 1989; Okabe and Hirokawa, 1990; Okabe et al., 1993; Sabry et al., 1995; Takeda et al., 1995; Takeda et al., 1994). The most likely explanation at that time was that cytoskeletal polymers were stationary and the moving form could be small oligomers or heterodimers. This raised another question in slow axonal transport of cytoskeletal proteins. That is whether they are transported as assembled polymers or free subunits. This question was debated over two decades and early studies favored the free subunits transport model.

A breakthrough in the field of slow axonal transport has come from observations of axonal transport of green fluorescent protein (GFP) tagged neurofilament proteins (Roy et al., 2000; Wang et al., 2000). The authors used cultured sympathetic neurons which contain relatively few neurofilaments and often show discontinuity of neurofilament arrays along their axons called gaps. Gaps enabled the observation of axonal transport of

GFP tagged neurofilaments without the need of photobleaching or photoactivation. The authors took time-lapse movies of these gaps with short time intervals (usually less than 5 seconds). Surprisingly, they observed that neurofilaments moved rapidly and

21 asynchronously. The peak velocity of the movement was as high as 3 μm/s. The movement was frequently interrupted by prolonged pauses and reversal of directions.

Recent computational modeling studies revealed that neurofilaments spend about 97% of their time pausing during their journal along the axon, resulting in the slow transport rate

(Trivedi et al., 2007). In these studies, the authors observed that GFP tagged neurofilaments moved in a predominantly filamentous form, which favored the polymer transport model.

The failure of previous observations of cytoskeletal protein transport is likely due to relatively long observation intervals (typically five minutes or more) since people expected to observe slow synchronous movements. In addition, the bleach of the protein was only partial in case of photobleaching. If the residual unbleached fluorescence in the bleached zone exceeded the fluorescence intensity of a single polymer, then movement of a single polymer through the bleached zone could not be identified. This issue was solved by gaps.

22

The form in which cytoskeletal proteins move remains controversial even after people accept that these proteins move in a rapid intermittent manner. In one of the early studies,

Terada et al. infected transgenic mice lacking axonal neurofilaments with a recombinant adenoviral vector encoding cMyc-tagged NFM. They observed cMyc-tagged NFM proteins in 10-nm filament-free axons suggesting that neurofilaments can be transported as monomers or small oligomers, which supported the subunit transport model (Terada et al., 1996). However, due to lack of neurofilament polymers in their study system, they could not rule out the possibility that neurofilaments can also be transported as polymers.

Several other studies in either neuronal cell lines or DRG neurons have reported movement of GFP tagged neurofilaments in a predominantly punctate form (Helfand et al., 2003; Theiss et al., 2005; Yabe et al., 1999). In addition, Prahlad et al. also observed fast transport of non-filamentous neurofilament protein particles along axonal microtubules in squid axoplasm (Prahlad et al., 2000). All of these studies suggest that neurofilaments can be transported as monomers or small oligomers.

In contrast, Galbraith et al. observed transport of polymerized neurofilaments in squid giant axon (Galbraith et al., 1999). Studies on cultured rat sympathetic and cortical

23 neurons transfected with GFP-tagged neurofilament proteins also showed that neurofilaments move in a predominantly filamentous form (Ackerley et al., 2003; Roy et al., 2000; Uchida and Brown, 2004; Wang and Brown, 2001; Wang et al., 2000). Until

2005, a study from the Brown lab showed without doubt that neurofilament polymers did move (Yan and Brown, 2005). In this study, the authors took advantage of the naturally occurring gaps of the neurofilament array along axons of cultured rat sympathetic neurons. Because these gaps lack neurofilaments, moving filaments can be observed in isolation in these regions. When the authors observed a neurofilament entering a gap, they quickly captured it by permeabilizing the axon before it exited the gap. They found that the moving neurofilament captured was indeed a single filament polymer by doing electron microscopy.

Despite the controversy of the forms in which cytoskeletal proteins are transported, we should note that subunit model and polymer model are not necessarily mutually exclusive. For example, in cultured sympathetic neuron studies, a small population of moving structures is in punctate form, which may represent unassembled filament subunits (Uchida and Brown, 2004). In neuronal cell line studies, initially punctate

24 structures were the dominant moving forms. However, following the development of the cell, movement of both punctate and filamentous structures was observed (Helfand et al.,

2003; Yabe et al., 2001a). It is likely that neurofilaments move in both subunit and polymer forms. The predominant moving form varies depending on the differentiation state of the cell and the punctate structures may represent motile precursors of neurofilament assembly.

1.2.4 Neurofilament phosphorylation and transport

In the field of slow axonal transport, neurofilament is one of the most well characterized cargoes. However, mechanisms regulating axonal transport of neurofilaments are poorly understood. Several studies have suggested that phosphorylation of the Lys-Ser-Pro

(KSP) repeats in NFM and NFH tail domains could regulate neurofilament transport.

KSP repeats in NFH can be classified into two categories, KSPXX and KSPXK (where the final X is any amino acid except lysine), which are regulated by different protein kinases. Several studies suggest that Cdk-5 (Cyclin-dependent kinase 5) phosphorylates

KSPXK repeats (Bajaj and Miller, 1997; Guidato et al., 1996; Hisanaga et al., 1993;

Shetty et al., 1993; Sun et al., 1996), while ERKs (extracellular signal–regulated kinases) 25 and GSK (glycogen synthetic kinase) preferentially phosphorylate KSPXX repeats (Pant and Veeranna, 1995; Roder and Ingram, 1991). Several studies have shown that phosphorylation of neurofilaments increased interactions between neurofilaments which may lead to a reduction in neurofilament movement (Yabe et al., 2001b).

Studies from the Nixon lab revealed that regional neurofilament accumulation was selectively associated with phosphorylation of a subset of KSP motifs on NFH and NFM subunits triggered by oligodendroglial signaling. The accumulation could be caused by decreased axonal transport of neurofilaments. The authors also showed that, in absence of

NFH, regional neurofilament accumulation was accompanied by increased levels of phosphorylated KSPs on NFM subunits. They also showed that in case of Shiverer mutant mice, in which this site-specific phosphorylation was selectively inhibited, regional neurofilament accumulation was reduced (Sanchez et al., 2000).

At the same time, the Miller lab reported that glutamate reduced axonal transport of neurofilaments associated with phosphorylation of the side arms via activation of the mitogen-activated protein kinase family members (Ackerley et al., 2000). Another study

26 from the same lab in 2003 reported that phosphorylation of NFH side arms regulated axonal transport of neurofilaments (Ackerley et al., 2003). In this study, they analyzed neurofilament transport rate in neurons expressing mutated NFH subunits, in which known NFH phosphorylation sites were mutated to preclude phosphorylation or mimic permanent phosphorylation. Neurofilament movement in neurons expressing the mimic permanent phosphorylation NFH constructs was reduced accompanied by an increased proportion of pausing time. In contrast, neurofilament movement in neurons expressing dephosphorylation mimic NFH constructs was accelerated. In addition, application of roscovitine, an inhibitor of Cdk5/p35, also accelerated neurofilament transport. These results provide strong evidence supporting that neurofilament phosphorylation may regulate neurofilament transport.

1.3 Molecular motors

Intracellular transport of materials is critical for cell growth, function and survival. It supplies the axon with newly synthesized proteins and lipids, and clears damaged or mis- folded proteins. The driving forces for intracellular cargo transport are molecular motors,

27 which are classified into three major groups: (1) kinesin superfamily motors; (2) dynein motors; (3) myosin superfamily motors (Hirokawa, 1998; Karki and Holzbaur, 1999;

Vale, 2003). Since microtubules are the major polarized cytoskeletal components in axons and dendrites, cargo transport is usually driven by microtubule based motors, kinesins and . Microtubules in axons and distal dendrites are plus ends distal, whereas in the proximal dendrites, the polarity is mixed. Majority of kinesins move cargoes toward the plus ends of microtubules, while dyneins moves cargoes toward the minus ends of microtubules. In the synaptic regions, are the major cytoskeletal components. As a result, transport of cargoes in this region is driven by actin-based motors, . The barbed ends (plus ends) of actin filaments point to the plasma membrane in the synaptic regions. Majority of myosins move toward the barbed ends of actin filaments.

1.3.1 Kinesin superfamily (KIF)

Kinesin superfamily motors (KIFs) were classified into 14 families based on phylogenetic analyses (Aizawa et al., 1992; Hirokawa et al., 2010; Hirokawa and

Takemura, 2005; Lawrence et al., 2004; Miki et al., 2001). Kinesin motors share a 28 common structure, which usually contains a conserved globular motor domain, a short neck region, a long stalk domain and a globular tail domain. The motor domain contains microtubule and ATP binding sites, which can hydrolyze ATP and generate mechanical force. The motor domain and the neck region together are called the head domain.

Kinesin motors can also be categorized into three major groups based on the location of the motor domain within the molecule. N-kinesins have the motor domain located at the

N-terminus of kinesin and usually move toward the plus ends of microtubules. C-kinesins have the motor domain located at the C-terminus of kinesin and usually move toward the minus ends of microtubules. M-kinesins (kinesin-13 family) have the motor domain located in the middle of the molecule and have microtubule-depolymerizing activity instead of motor activity.

29

Figure 1.5 Structures of the major kinesins Generally, kinesins consist of a motor domain and a coiled-coil domain. Kinesins can be categorized into N-kinesins, M-kinesins and C-kinesins, based on the location of the motor domain. Only the kinesin 13 family (KIF2) contains M-kinesins and only the kinesin 14A and kinesin 14B families contain C-kinesins (KIFC2 and KIFC3). All other families consist of N-kinesins. (Hirokawa et al., 2009) 30

Kinesin-1 (conventional kinesin, KIF5) was the first identified kinesin member. It is a heterotetramer composed of two heavy chains and two light chains (Hirokawa et al.,

2009). The motility of kinesin-1 has been extensively studied. When kinesin-1 encounters microtubules, it usually moves several hundred steps (about several microns) before detaching from microtubule tracks, which is termed processive movement (Block et al.,

1990; Hackney, 1995; Howard et al., 1989; Vale et al., 1996). A commonly accepted model accounting for this processive movement is the coordinated hand-over-hand model, in which two heads of kinesin-1 bind alternatively while it moves along microtubule (Vale and Milligan, 2000).

The head domain, which includes the motor domain and the neck region, is critical for the processive movement of kinesin-1 (Hancock and Howard, 1998; Young et al., 1998).

Several studies have shown that the neck region plays an important role in regulating the run length (distance traveled per microtubule encounter) of the motor. By adding positive charge to the neck, the run length of kinesin mutants was increased by four fold. In contrast, by adding negative charge to the neck, the run length was decreased (Thorn et al., 2000).

31

The tail domain of kinesin-1 is involved in cargo binding and interacting with kinesin light chains (Brady, 1995). In mammals, there are two kinesin-1 light chain , KLC1 and KLC2 (Rahman et al., 1998). The tail domain is also involved in auto-inhibition of the motor. The tail domain blocks the initial interaction of the head domain with microtubule by folding back and interacting with the head domain directly (Cai et al.,

2007; Dietrich et al., 2008). In contrast, kinesin light chains suppress the interactions between the tail and the head domains, resulting in promoting activation of kinesin-1 for cargo transport (Wong and Rice, 2010).

Kinesin and neurofilament transport

Several studies have suggested that kinesin-1 may be the anterograde motor for neurofilaments. Shea and colleagues proposed that kinesin-1 is a neurofilament motor based on the observation that anti-kinesin-1 antibody blocked transport of neurofilament dots (Yabe et al., 1999) and the interaction study of kinesin-1 and neurofilament (Jung et al., 2005; Yabe et al., 2000), although there are several concerns of these studies (See

Chapter 4). In mammals, there are three kinesin heavy chain genes, kinesin-1A, 1B and

1C. Kinesin 1A and 1C are neuron specific, whereas kinesin-1B is expressed ubiquitously 32

(Xia et al., 1998). In these studies, the authors did not specify which isoform of kinesin-

1s was principle for neurofilament transport.

Studies from kinesin-1A knockout mice have provided us some clue that kinesin-1A may be a principle an anterograde motor for neurofilaments in axons (Xia et al., 2003). In support of this proposal, the Brown lab observed a 75% reduction in the frequency of neurofilament movement in cultured neurons from kinesin-1A knockout mice, which could be rescued by kinesin-1A. In addition, headless kinesin-1A and kinesin-1C each inhibited neurofilament transport in both directions in a dominant-negative manner in wild type neurons (Uchida et al., 2009). These data suggest that kinesin-1A is the principal anterograde motor for neurofilaments.

1.3.2 Dynein/dynactin complex

Dyneins are microtubule minus end directed motors. Unlike kinesins, which have 14 families, there are only two groups of dyneins, axonemal and cytoplasmic dyneins (Hook and Vallee, 2006). Axonemal dyneins are involved in ciliary and flagellar beating, while cytoplasmic dyneins are responsible for of cargoes, mitosis and cell 33 movement. Cytoplasmic dyneins consist of cytoplasmic dynein 1 and cytoplasmic dynein

2, which are distributed differently and have different functions. Cytoplasmic dynein 1 is the most abundant dynein expressed in all microtubule containing cells. It is involved in retrograde transport of lysosomes, endosomes etc. It is critical for maintaining normal cell function. Expression of cytoplasmic dynein 2 is limited in cilia and flagella, where it is involved in intraflagellar transport for axonemal maintenance (Mikami et al., 2002;

Pazour et al., 1999; Porter et al., 1999).

Dynein motors are multi-protein complex consist of two dynein heavy chains (DHCs), which are homodimers, and several dynein intermediate, light intermediate and light chains (Karki and Holzbaur, 1999; Pfister et al., 2006). The molecular weight of DHC is more than 500 kDa. The motor domain of DHC is about 380 kDa, which consists of six

AAA ATPase units, which hydrolyze ATP to generate mechanical force (Samso et al.,

1998) (Neuwald et al., 1999). Microtubule binding site of DHC is located downstream of the ATPase region in the stalk domain (Burgess et al., 2003; Gee et al., 1997). The rest

160 kDa NH2-terminal domain forms the base of the molecule, where accessory subunits bind.

34

Cytoplasmic dynein does not appear to be a processive motor without its activator, dynactin, which is also a large multi-protein complex consisting of 11 subunits. Dynactin facilitates dynein-driven transport in two ways: (1) it functions as an adaptor linking dynein to cargoes; (2) by binding to microtubule, it increases the processivity of dynein

(King and Schroer, 2000). Dynactin consists of two distinct domains (Figure 1.6): a cargo-binding domain and a projecting arm which binds microtubule. These two parts are linked by a triangular shoulder. The major components of the cargo-binding domain are eight subunits of the actin-related protein (Arp1). The projecting arm contains the N- terminal third of the p150Glued dimer, which is a vertebrate homologue of

Glued gene product. It is the largest subunit of dynactin complex, which is encoded by one gene yielding at least three alternative isoforms. The shoulder comprises the C- terminal two-thirds of the p150Glued dimer, four dynamitins (p50), and two p24s. Over- expression of dynamitin disrupts the function of dynactin by releasing p150Glued from

Arp1 (Burkhardt et al., 1997).

35

Figure 1.6 Schematic structure of dynein/dynactin Dynein itself is a complex of heavy (HC), intermediate (IC) and light chains. It interacts with the p150glued subunit of the dynactin complex through its intermediate chains (arrow). Actin-related proteins (Arp1s) are the major components of the cargo binding domain of dynactin. (Schliwa and Woehlke, 2003)

Dynein/dynactin complex and neurofilament transport

Neurofilaments are transported bi-directionally along axons. Several studies including studies from the Brown lab suggest that dynein/dynactin complex is the retrograde motor responsible for neurofilament transport. Studies from He et al. showed that knock down of dynein heavy chain (DHC) expression using siRNA significantly decreased retrograde 36 neurofilament transport (He et al., 2005). In this study, neurofilament transport in anterograde direction was increased. Studies from the Brown lab using siRNA to knock down DHC confirmed the decrease in retrograde neurofilament transport but instead of an increase in anterograde neurofilament transport, the authors observed a decrease in anterograde neurofilament movement. Functional blocking antibodies to dynactin also decreased neurofilament transport in both directions. In addition to that, using dominant negative approaches, for example over-expression of dynamitin which disrupts dynactin complex, they also observed a reduction in neurofilament transport in both directions

(Uchida et al., 2009). These results suggest that motors driving bi-directional neurofilament transport may be functional interdependent.

In addition to live-cell imaging studies, there are also studies showing interactions between neurofilament and dynein/dynactin. Shah et al. discovered that dynein/dynactin co-purified with neurofilaments. In addition, using immuno-electron microscopy, the author observed decoration of neurofilaments with anti-dynein antibody (Shah et al.,

2000). Later, studies from Wagner et al. identified a direct interaction between NFM subunit and dynein intermediate chain using yeast two hybrid and affinity

37 chromatography assays. Using AFM, they observed association of neurofilaments with the base of dynein, suggesting that neurofilament is a cargo for dynein (Wagner et al.,

2004). If dynein intermediate chain interacts primarily with NFM, it would be interesting to investigate whether there are neurofilaments moving retrogradely along axons in case of NFM knockout.

1.3.3 Myosin superfamily

Long distance transport of proteins in cells is usually driven by microtubule-based motors, kinesins and dyneins. However, there are still regions of cells which lack microtubules. Protein transport in these regions is usually driven by actin-based motor, myosins, which are also ATP-dependent motors. A myosin heavy chain typically contains a head, neck and tail domain (Figure 1.7). The head domain is responsible for actin binding and ATP hydrolysis to generate mechanical force. The core structure of the head domains of myosins are relatively conserved across species with the exception of several surface loops and the amino-terminus (Sellers, 2000). The neck domain serves as a linker to transduce force and as a binding site for myosin light chains. The tail domain is involved in cargo binding as well as dimerization of heavy chains. Myosin motors 38 generally move toward the barbed (+) ends of actin filaments except myosin VI which moves toward the (-) ends. Myosin superfamily motors have been divided into 17 classes based on phylogenetic analysis of the motor domains.

Figure 1.7 Structures of three major myosin proteins Myosin family proteins consist of head, neck and tail domains. The head domain binds actin filaments and has ATPase activity. The neck domain interacts with light chains, which regulate the head domain. The neck regions of myosin I and myosin V interact with light chains, whereas the neck domain of myosin II interacts with essential light chain and regulatory light chain. These light chains are differently regulated by Calcium. The tail domain determines the specific role of each myosin in the cell (Harvey Lodish, 2000).

39

Myosin II is conventional myosin, which is mainly involved in muscle contraction. It also exists at pre- and post-synaptic regions of neurons, where it is involved in neurotransmitter release at pre-synaptic region (Mochida et al., 1994) and regulating the structure and function of dendritic spines at post-synaptic region (Ryu et al., 2006).

Recently, Myosin II has been shown to play an important role in neuronal migration.

Studies from Solecki et al. showed that myosin II and F-actin dynamics drive the coordinated movement of the and soma during glial-guided neuronal migration (Solecki et al., 2009).

Myosin V is an unconventional myosin motor, which is functional as a dimer. Mutations in the myosin-Va gene are associated with Griscelli disease (Pastural et al., 1997). Tabb et al. showed that ER vesicles were transported on actin filaments by myosin V and proposed a dual filament model of vesicle transport (Tabb et al., 1998). Later, studies from Huang et al. showed that myosin Va interacted directly with a microtubule-based motor, KhcU (Huang et al., 1999), suggesting that interaction of different motor molecules may coordinate intracellular transport. A study of myosin Va movements in normal and dilute-lethal (myosin Va-/-) (Mercer et al., 1991) axons showed that myosin

40

Va activity was not necessary for long-range transport along axons, but it was necessary for local transport in regions that lack microtubules, such as presynaptic terminals

(Bridgman, 1999).

Myosin and neurofilament transport

Neurofilament transport mainly depends on the integrity of microtubules in the axon.

Actin filaments are not the major route for neurofilament transport in the axon. Minor impairments of neurofilament movement were observed in axons of cultured sympathetic neurons treated with latrunculin, which disrupts actin filaments (Francis et al., 2005), and in neurites of NB2a/d1 cells treated with cytochalasin B, which also disrupted filamentous actin (Jung et al., 2004), suggesting that actin filaments play a minor role in neurofilament transport. In support of this idea, Jung et al. found that neurofilament translocation was perturbed by inhibition of actin-based myosin motors with either ML-7

( kinase inhibitor) or 2,3-butanedione-2-monoxime (myosin ATPase inhibitor) (Jung et al., 2004). In addition, Nixon and colleagues demonstrated an interaction between myosin Va and neurofilaments, which is important in establishing the normal distribution of myosin Va and neurofilament density in axons (Rao et al., 2002a). 41

Based on these observations, recent studies from the Brown lab conducted by Alami et al. showed that the pausing time of neurofilaments was increased in cultured neurons from dilute lethal mice which lack myosin Va. The authors proposed that myosin Va could enhance the efficiency of neurofilament movement by delivering neurofilaments to microtubules to reduce the pausing time of neurofilaments (Alami et al., 2009). However, this proposed mechanism is called into question by a recent finding from the Nixon lab showing that NFL interacts with the motor domain instead of the tail domain of myosin

Va, suggesting that neurofilaments are unlikely to be the cargo for myosin Va. These authors also showed that neurofilaments are unlikely to be the rails for myosin Va based on the observation that NFL deletion had no effect on the fast transport of vesicular organelles propelled by myosin Va (Rao et al., 2011). Since it has been show that myosin

Va can interact directly with a microtubule-based motor, KhcU (Huang et al., 1999), it is possible that myosin Va may exert its effect on neurofilament transports through interaction with kinesin motors instead of through actin-based transport system.

42

In this thesis, I investigated the effect of a kinesin-1A mutant on neurofilament transport.

This same mutation causes hereditary spastic paraplegia in patients. In addition, I also characterized the interaction between kinesin-1A/dynein and neurofilaments.

1.4 Neurofilaments and neurodegenerative diseases

Mutations in neurofilament genes have been associated with several neurodegenerative diseases. In addition, abnormal accumulations of neurofilaments are a pathological hallmark of many neurodegenerative diseases, including Amyotrophic Lateral Sclerosis

(ALS), Alzheimer’s disease, Parkinson’s Disease, Charcot-Marie-Tooth (CMT), giant axonal neuropathy (GAN) and diabetic neuropathy. The accumulation can be caused by alteration of neurofilament gene expression, neurofilament gene mutations, deficient axonal transport or abnormal post-translational modifications and proteolysis. In this section, I will review some of these diseases, in which involvement of neurofilament has been extensively studied.

43

1.4.1 Amyotrophic Lateral Sclerosis

Amyotrophic Lateral Sclerosis (Lou Gehrig’s disease) is a late-onset progressive motor neuron disease. A pathological hallmark of ALS is intraneuronal aggregates of neurofilaments, termed spheroids. About 90% of ALS cases are sporadic, whereas the other 10% are familial cases (Averback, 1981; Carpenter, 1968; Delisle and Carpenter,

1984; Hirano et al., 1984; Ince et al., 2011; Munoz et al., 1988; Murayama et al., 1992).

Although the mechanisms leading to neurofilament accumulation in ALS are not well understood, over-expression of human neurofilament subunit provokes massive accumulations of neurofilaments and severe skeletal muscle atrophy similar to those found in ALS (Cote et al., 1993; Wong et al., 1995; Xu et al., 1993). Remarkably, studies from Julien and colleagues showed that over-expression of human NFL (hNFL) could rescue the motor neuron disease caused by excess expression of hNFH proteins in a dosage-dependent manner. The additional hNFL reduced perikaryal swellings and restored axonal radial growth (Meier et al., 1999). Lee et al. discovered that expression of an NFL point mutant caused abnormal accumulations of neurofilaments accompanied by selective degeneration of spinal motor neurons and severe skeletal muscle atrophy, 44 suggesting that aberrant neurofilament accumulation could contribute to neuronal death

(Lee et al., 1994).

Mutations in copper/zinc superoxide dismutase gene (SOD1) account for approximately

20% of familial ALS cases. Transgenic mice expressing mutant SOD1 displayed neurofilament accumulation (Borchelt et al., 1998; Tu et al., 1996) and developed a motor neuron disease similar to mice over-expressing NFL or hNFH. Impairment of axonal transport by neurofilament inclusions was observed in motor neurons during the onset and progression of motor neuron disease in transgenic mice expressing human

SOD1 with a G93A mutation (Zhang et al., 1997). Additionally, another two studies also revealed that axonal transport was disrupted in mice with SOD1 G37R and G85R mutations (Borchelt et al., 1998; Williamson and Cleveland, 1999). To determine the involvement of neurofilaments in SOD1-mediated diseases, SOD1 mutant mice were crossed with transgenic mice with altered neurofilament contents. Williamson et al. found that the onset and progression of disease caused by G85R SOD1 mutation were significantly slowed by removing NFL (Williamson et al., 1998).

45

Surprisingly, over-expression of mouse NFL or mouse NFH in SOD1 G93A mice (Kong and Xu, 2000) and hNFH in SOD1 G37R mice (Couillard-Despres et al., 1998) had a protective effect. The life spans of the animals were increased by 15% and 65% respectively. Since neurofilaments have multiple calcium binding sites and phosphorylation sites, it is possible that the perikaryal neurofilament aggregates function as calcium chelators (Roy et al., 1998) or a phosphorylation sink (Nguyen et al., 2001) to reduce calcium toxicity or the detrimental hyperphosphorylation of tau and other neuronal substrates by CDK5, although the phosphorylation sink mechanism was called into questions by a recent study (Lobsiger et al., 2005). In this study, the authors generated SOD1G37R mice in which the tail domains of NFM and NFH were removed.

The tail domains of NFM and NFH are usually heavily phosphorylated and were previously proposed as phosphorylation sinks. Deleting the tail domains of NFM and

NFH delayed disease onset and extended survival in mutant SOD1 mice instead of accelerating the disease, suggesting that the phosphorylation sink mechanism is unlikely to be an essential explanation for the benefit of altered neurofilament subunit composition on SOD1 mutant mice. It seems that the benefit may primarily be an axonal one, in which reduction of neurofilament content or removal of the tail domains of NFM and NFH

46 provides a more flexible axoplasm, which enhance the movement of slow axonal transport cargoes since disruption of axonal transport is usually associated with the onset and progression the disease. This may also explain the benefit of removing NFL in case of Williamson’s study.

1.4.2 Charcot-Marie-Tooth

Charcot-Marie-Tooth are the most common inherited disorders that affect the peripheral nerves (Skre, 1974). They are characterized by loss of muscle tissue and touch sensation predominantly in the feet and legs. CMT are classified into several categories, including

CMT1, CMT2, CMT3, CMT4 and CMTX. CMT2 is autosomal dominant axonal form of the diseases. Accumulation of neurofilaments was first discovered in CMT2 (Vogel et al.,

1985). Mutations of NEFL gene, which encodes NFL subunit, are associated with

CMT2E and CMT1F. An A998C transversion was discovered in the first exon of NEFL, which converts a conserved Gln333 amino acid to proline in a large Russian family with

CMT2 (Mersiyanova et al., 2000). Interestingly, expression of NFLQ333P in cultured neurons disrupted neurofilament assembly and transport as well as the localization of mitochondria (Brownlees et al., 2002). This mutant also affected bi-directional fast 47 axonal transport, providing possible mechanisms by which this mutant could be involved in CMT pathogenesis (Perez-Olle et al., 2005).

Another target of several NFL mutations is Pro at codon 22. Georgiou et al. identified a

Pro22Ser NFL mutation in a large Slovenian family (Georgiou et al., 2002). The same mutation was also identified in an Italian family. Nerve biopsy examinations in these patients revealed an axonopathy characterised by giant axons composed of disorganized neurofilaments (Fabrizi et al., 2004). Studies from cultured cells indicated that the Pro22

CMT mutation caused abnormal neurofilament accumulation by disrupting oligomer formation and the aggregates were mitigated by phosphorylation with PKA, which makes it a potential therapeutic target (Sasaki et al., 2006).

1.4.3 Giant axonal neuropathy

Giant axonal neuropathy (GAN) is an autosomal recessive disease in which both the peripheral and the central nervous system are affected. It is caused by mutations in the gene GAN, which encodes the protein named (Bomont et al., 2000). A characteristic feature of GAN is accumulation of intermediate filaments, especially 48 neurofilaments, leading to segmental distension of the axons. Axon swellings and spheroids piled with neurofilaments were observed in the spinal cord, cerebral cortex and sural nerve of GAN patients (Boltshauser et al., 1977; Peiffer et al., 1977). The minimum separations between neurofilaments in GAN were 12-30 nm compared with 24-60 nm in controls and the average minimum neurofilament diameter was 12.4 nm in GAN compared with 10.1 nm in controls (Donaghy et al., 1988), suggest that the side arms of

NFM and NFH may not normally protrude from the backbone of the filament. It is possible that instead of protruding to form the cross-bridges, the side arms lie longitudinally along the filaments, resulting in larger neurofilament diameter and smaller spaces between neurofilaments.

In a rat model of GAN, axonal transport of neurofilaments was accelerated resulting in a reduction of neurofilaments proximally and an increase of neurofilaments distally

(Monaco et al., 1985). To study how disruption of gigaxonin's function leads to neurodegeneration, Ding et al. generated GAN null mice. The neurological phenotypes and pathological lesions developed in GAN-/- mice were reminiscent of human GAN.

Disruption of gigaxonin leads to accumulation of microtubule-associated protein (MAP)

49

8, which in turn may cause axonal transport impairment (Ding et al., 2006). This group also discovered a direct interaction between gigaxonin and MAP1B light chain. This interaction was shown to enhance the stability of microtubules, which serve as rails for axonal transport (Ding et al., 2002). In addition, they identified an interaction between gigaxonin and folding cofactor B (TBCB), which resulted in degradation of

TBCB through ubiquitin-proteasome pathway. Accumulation of TBCB caused by disruption of gigaxonin may be a causative factor of cytoskeletal pathology in GAN

(Wang et al., 2005).

1.4.4 Diabetic neuropathy

Diabetic neuropathy is nerve damage caused by high blood glucose level. It is characterized by axonal atrophy, impaired axonal transport and slowing of conduction velocity. All these features depend on axonal cytoskeleton integrity and especially on the neurofilament network. In agreement with this, abnormalities of neurofilaments have been identified in models of diabetes. Rats with streptozotocin-induced diabetes showed impairment of neurofilament transport resulting in a proximal increase and a distal decrease of axonal cross-sectional area (Medori et al., 1985). The BioBreeding (BB) rats, 50 which are another model for diabetes, showed similar changes in slow axonal transport of neurofilament and axonal size. Notably, transport was not impaired in BB rats maintained the normal amount of blood glucose (Medori et al., 1988). Accumulations of highly phosphorylated NFH were observed in sympathetic autonomic ganglia of diabetic patients (Schmidt et al., 1997). In addition, Fernyhough et al. observed a significant increase in phosphorylation of JNK in DRG and sural nerve, which correlated with elevated neurofilament phosphorylation in two animal models of type 1 diabetes

(Fernyhough et al., 1999). Diabetic rats showed reductions of mRNA expression of neurofilament subunits and α-tubulin in sensory neurons accompanied reductions of cytoskeletal proteins in distal axons resulting in a decrease of conduction velocity and axon atrophy (Scott et al., 1999). These findings suggest that neurofilament abnormalities may contribute to the development of diabetic neuropathy or may be a result of this disease.

To directly test the relationship between neurofilaments and diabetic neuropathy,

Zochodne et al. superimposed streptozotocin-generated diabetes on NFH-beta- galactosidase transgenic mice, in which the C-terminus of NFH was replaced by beta-

51 galactosidase. The fusion proteins aggregated in perikarya of neurons resulting in depletion of neurofilaments from axons. Despite similar levels of hyperglycaemia, NFH- beta-galactosidase diabetic mice developed progressive reduction of conduction velocity in their motor and sensory fibers earlier than diabetic mice with normal neurofilaments.

In addition, NFH-beta-galactosidase diabetic mice displayed increased axonal atrophy

(Zochodne et al., 2004). These results indicate that neurofilaments in axons may have a protective effect against diabetic neuropathy.

1.5 Axonal transport and neurodegenerative diseases

Kinesin mutations in neurodegenerative diseases

Defects in kinesin motors could cause synaptic defects or axonal dieback due to reduced supply of new proteins and lipids from the cell body to the distal part of the axon.

Relatively few degenerative diseases have been directly linked to mutations in kinesin motors, which is probably due to functional redundancy in the kinesin superfamily.

Another possibility is that mutations in kinesin motors may be lethal.

52

Recent studies identified several mutations in kinesin-1A in families with pure and complex hereditary spastic paraplegias. Hereditary spastic paraplegias are a group of neurological disorders characterized by progressive distal degeneration of the longest ascending and descending axons in the spinal cord, leading to lower limb spasticity and weakness. One of the dominantly inherited forms of this disease (spastic gait type 10, or

SPG10) is caused by point mutations in kinesin-1A.

In addition, a Q98L missense mutation in the ATP binding consensus of motor domain in kinesin-3Bβ (KIF1Bβ) has been associated with Charcot-Marie-Tooth disease type 2A

(axonal types of degeneration predominating). The mutant kinesin-3Bβ has decreased

ATPase activity and motility, suggesting that disruption of axonal transport caused by a haploinsufficiency of this motor may underlie human peripheral neuropathy (Zhao et al.,

2001).

Dynactin mutations in neurodegenerative diseases

Mutations in DCTN1 gene, which encodes p150Glued, have been identified in at least two distinct forms of neurodegenerative diseases, distal spinal and bulbar muscular atrophy 53

(dSMA) (Puls et al., 2003) and Perry syndrome (Farrer et al., 2009). Distal spinal and bulbar muscular atrophy is a late-onset autosomal dominant lower motor neuron disease with prominent bulbar symptoms. The mutation causing dSMA (G59S) occurs within the conserved glycine-rich (CAP-Gly) domain of p150Glued, inhibiting the interaction between dynactin and microtubules (Levy et al., 2006). An autopsy study revealed motor neuron degeneration and loss of axons in the ventral horn of the spinal cord and hypoglossal nucleus of the medulla in a patient with dSMA. In addition, abnormal accumulation of dynein and dynactin in motor neurons was revealed by immunohistochemistry study (Puls et al., 2005).

Perry syndrome is a progressive brain disease that is characterized by Parkinsonism, depression, weight loss, and hypoventilation. Five disease-segregating mutations within the CAP-Gly domain of p150Glued have been identified in eight families (Farrer et al.,

2009). The mutations cluster in or near the surface-exposed microtubule-binding motif compared to the dSMA mutation which is buried within the folded domain of p150Glued.

These mutations also disrupt the interaction between dynactin and microtubule probably

54 through a different mechanism which potentially accounting for the selective vulnerability of distinct neuronal populations in dSMA and Perry syndrome.

55

Chapter 2 Materials and Methods

Methods of astroglial and cortical cells co-culture, cloning of cMyc-N256S-kinesin-1A and cMyc-wild type-kinesin-1A, live cell imaging and analysis, electroporation and immunocytochemistry were reproduced from Wang and Brown 2010 paper (Wang and

Brown, 2010) with revisions. The original paper was written by both authors.

2.1 Cell culture

2.1.1 Astroglial cells and cortical neurons co-culture

Cortical neurons were cultured using the glial sandwich technique of Banker (Kaech and

Banker, 2006). Wild type ICR pregnant female mice were obtained from Harlan

Laboratories (Indianapolis, IN). To prepare glial cultures, the cerebral cortices from 3-5 postnatal day 0 mouse pups were dissociated in phosphate buffered saline (PBS;

Invitrogen, Carlsbad, CA) containing 0.25% [w/v] trypsin (Worthington Biochemical

Corp., Lakewood, NJ), 1% [w/v] DNase-I (Sigma, St. Louis, MO) and 0.54mM EDTA

(Sigma) for 15 minutes. The activity of trypsin was terminated by glial medium, which 56 consisted of Minimum Essential Medium (Invitrogen) supplemented with 10% [v/v] horse serum (Invitrogen), 0.7% [w/v] glucose (Sigma) and 16 μg/ml gentamicin

(Invitrogen). After pipetting with a glass Pasteur pipette for approximately 20 times, cells were filtered through a cell strainer (BD Falcon, Bedford, MA). Each brain usually yielded 8*106 to 10*106 cells. Cells were diluted to the desired concentration in glial medium and plated onto T75 flasks (BD Falcon) at a density of 4*105 – 5*105 cells per

ml (11 ml total) and cultured at 37°C/5% CO2. Cells were fed every three days. Three days before cortical neuron culture, glial cells were passaged and cultured on acid washed glass coverslips with paraffin wax dots. One day before cortical neuron culture, glial medium was replaced by plating medium, which consisted of Neurobasal medium

(Invitrogen) supplemented with 2% [v/v] B-27 Supplement Mixture (Invitrogen), 0.27%

[w/v] glucose, 2 mM glutamine (Invitrogen), 37.5 mM NaCl (Sigma), 5% [v/v] fetal bovine serum (FBS; Thermo Scientific, Waltham, MA), 16 μg/ml gentamicin, and 2.5

μM cytosine arabinoside (AraC; Sigma). Glial cells could be cryopreserved in liquid nitrogen for long term storage.

57

To prepare neuronal cultures, the cerebral cortex of one postnatal day 0 mouse pup was dissociated in PBS containing 0.025% [w/v] Trypsin, 0.27mM EDTA and 0.5% [w/v]

DNase-I. Cells were spun down by centrifuging at 100 g for 5 minutes at room temperature. After electroporation, cells were plated onto glass-bottomed dishes that had been coated with poly-D-lysine (Sigma) and laminin (BD Biosciences, San Jose, CA) at a concentration of around 1*105 cells per ml in plating medium. Two hours after plating, plating medium was replaced completely with plating medium in which glial cells were cultured. Glass coverslips bearing glia (>80% confluency) were suspended over the neurons using dots of paraffin wax as spacers, and the resulting sandwich cultures were maintained at 37°C/5% CO2.

After two days, plating medium was replaced by culturing medium, which was identical to plating medium except that it lacked serum. Every four days, half of the medium was removed and replaced with fresh medium.

58

2.1.2 HEK 293T cell culture

HEK 293T cells were cultured in DMEM medium (Invitrogen) containing 10% [v/v] FBS

and 16 μg/ml gentamicin in 60 mm plastic dishes (BD Falcon) at 37°C/5% CO2. Cells were passaged when they grew to nearly 100% confluency. To passage the cells, the cells were washed twice with warm PBS and then dissociated from the dishes using PBS containing 0.25% [w/v] trypsin and 0.54mM EDTA. Cells were spun down by centrifuging at 1100 rpm for 5 minutes using a Sorvall RT 6000D centrifuge (GMI Inc.,

Ramsey, MN) and then plated at a desired concentration onto 60 mm dishes.

2.2 Cloning

cMyc-N256S-kinesin-1A and cMyc-wild type-kinesin-1A

Mouse kinesin-1A cDNA (Genbank accession No.BC058396, I.M.A.G.E. clone

6824963) was obtained from American Type Culture Collection (Manassas, VA) and then subcloned into pEGFP-C1 (Clontech, Mountain View, CA) lacking the EGFP sequence, as previously described (Uchida et al., 2009). Briefly, the kinesin-1A cDNA clone was excised from the parent pYX-Asc vector with BspEI and BamHI and then 59 subcloned into the pEGFP-C1 vector that had been digested with AgeI and BamHI to excise the EGFP sequence. The N256S-kinesin-1A plasmid construct was generated using a QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA) with forward primer 5’-GGC AAA GAA TAT CAG CAA GTC GCT GTC GGC CC and reverse primer 5’-GGG CCG ACA GCG ACT TGC TGA TAT TCT TTG CC. The wild type and mutant kinesin-1A constructs were tagged with cMyc at their N-terminus with forward primer 5’-CTA GCT CCG GAA TGG AGC AGA AGC TGA TCA GCG AGG

AGG ACC TGG AG and reverse primer 5’-TCG ACT CCA GGT CCT CCT CGC TGA

TCA GCT TCT GCT CCA TTC CGG AG. The resulting cMyc-N256S-kinesin-1A and cMyc-kinesin-1A constructs were confirmed by DNA sequencing of their open reading frames.

EGFP-mNFM

The EGFP-mNFM plasmid construct, which encodes the codon-optimized F64L/S65T variant of green fluorescent protein attached to the N-terminus of mouse neurofilament protein M, the cDNA (GenBank accession DQ201636) of which was obtained by reverse

60 transcription-polymerase chain reaction (RT-PCR) using RNA from wild-type P24 mouse cerebellum, was described by Yan et al. (Yan et al., 2007).

61

EB1-YFP

The EB1-YFP construct was a general gift from Dr. Chen Gu at the Ohio State

University. It was originally constructed by inserting the coding sequence of EB1 into pEYFP-N1 between BglII and HindIII sites (Gu et al., 2006).

pNFL / pNFM / pNFH (mouse) pNFL and pNFM were previously constructed by Dr. Yan in the Brown lab (Yan et al.,

2007). Briefly, mouse neurofilament protein L and M cDNAs were obtained by reverse transcriptase-polymerase chain reaction (RT-PCR) using RNAs from wild type P24 mouse cerebellum (Genbank Accession Numbers DQ201635 and DQ201636, respectively). For NFM, BamHI sites were introduced at both ends of the amplified

DNA. The PCR product was subcloned into BamHI site of pEGFP-C1 vector. The EGFP sequence was then removed using AgeI and BspEI from the plasmid. The sticky ends were re-ligated. For NFL, Xho I and EcoR I sites were introduced at ends of the amplified DNA. The PCR product was subcloned into Xho I and EcoR I sites of pEGFP-

C1 vector. The EGFP sequence was then removed using AgeI and BspEI from the

62 plasmid. pNFH was constructed by Dr. Trivedi in the Brown lab using Don Cleveland’s

NFH clone as a template.

mCherry-vimentin and EGFP-kinesin-5 mCherry-vimentin was constructed by Dr. Colakoglu in the Brown lab (Colakoglu and

Brown, 2009). EGFP-vimentin with human vimentin cDNA was provided by Robert

Goldman. The EGFP was replaced with mCherry using NheI and BsrG1 sites. The mCherry was cut from Cherry-C1 vector using the same pair of enzymes. EGFP-kinesin-

5 was a general gift from Peter Baas’ lab (Myers and Baas, 2007).

HA-tagged headless kinesin-1A

The hemagglutinin (HA)-tagged headless kinesin-1A construct was created by amplifying N-terminally truncated portions of the corresponding cDNA with PCR and subcloned into the pCGN-HA mammalian expression vector of Tanaka and Herr (Tanaka and Herr, 1990).

63

2.3 Transfection

2.3.1 Electroporation

The dissociated cortical neurons were transfected by electroporation prior to plating using an Amaxa Nucleofector™ (Lonza Inc., Walkersville, MD) with the mouse neuron nucleofection kit (VPG-1001) and program O-05. The volume of the cell suspension was

100 μl and the cell density ranged from 4*106 to 6*106 cells per ml. For the experiments on neurofilament movement and distribution, I used 2 μg EGFP-mNFM construct either alone or in addition to 2 μg cMyc-N256S-kinesin-1A or 2 μg cMyc-wild type-kinesin-1A construct. For microtubule polarity experiments, 2 μg EB1-YFP construct was used.

2.3.2 Lipofection

HEK cells were re-plated one day before lipofection. On the day of lipofection, cells reached 80% confluency in 60 mm culture dishes. I used Lipofectamine LTX and Plus reagent for transfection (Invitrogen). 3 μg DNA of each construct was mixed in 250 μl

DMEM with 40 μl Plus reagent and incubated for 15 minutes. During this time, I changed the medium of the cells to 2 ml DMEM without serum. After 15 minutes, I 64 added 250 μl DMEM containing 10 μl Lipofectamine to the Plus mixture, and then incubated for another 15 minutes. After incubation, the mixture was added to HEK cells.

After 3 hour, the medium was changed back to normal DMEM medium containing FBS.

Cells were cultured at 37°C/5% CO2 overnight before used for immunoprecipitation.

2.4 Live cell imaging and analysis

2.4.1 Imaging of neurofilament movement through gaps

To image neurofilament movement, cortical neurons were observed after 8 to 12 days in culture by epifluorescence microscopy on a Nikon TE300 inverted microscope (Nikon,

Garden City, NY) using a 100x Plan Apo VC 1.4NA oil immersion objective. The observation medium consisted of Hibernate-E (BrainBits, Springfield, IL) supplemented with 2% [v/v] B27 Supplement Mixture, 0.3% [w/v] glucose, 1 mM L-glutamine, 37.5 mM NaCl, and 10 μg/ml gentamicin. The temperature on the microscope stage was maintained using an Air Stream incubator (Nevtek, Williamsville, VA). A layer of dimethylpolysiloxane fluid (Sigma, 5 centistokes) was floated over the observation medium to prevent evaporation. For time-lapse imaging, the exciting light from the 65 mercury arc lamp was attenuated 12-fold using neutral density filters, and images were acquired with one second exposures at four second intervals using a Micromax 512BFT cooled CCD camera (Roper Scientific, Trenton, NJ) and MetaMorph™ software

(Molecular Devices, Sunnyvale, CA). All movies were 15 minutes in length. It was necessary to adjust the focus occasionally during movie acquisition to correct for focus drift.

2.4.2 Analysis of neurofilament movement

Neurofilament movement was analyzed by tracking the position of the leading or trailing ends of the filaments in successive frames of the time-lapse image movies using

MetaMorph™ software. All objects greater than or equal to 10 pixels (1.31 μm) in length were analyzed if they moved a total distance of at least 50 pixels (6.55 μm) and could be tracked through at least three successive frames of the movie.

To calculate the frequency of movement, I classified each neurofilament as anterograde or retrograde based on its preferred direction of motion and then counted the number of anterograde and retrograde moving filaments per movie. Thus for each movie I obtained 66 two frequency measurements, one anterograde and one retrograde. Ninety seven percent of the filaments exhibited a preferred direction of movement, defined as moving at least

70% of their time in the same direction. The remaining 3% of the filaments, which spent more than 30% of the time moving in the opposite direction, were each considered to represent separate anterograde and retrograde moving events.

To calculate the flux, I grouped all the filaments in each 15-minute movie together and measured the total anterograde and retrograde distance moved. Thus for each movie I obtained two flux measurements, one anterograde and one retrograde.

For the analysis of bout velocity, bout duration and bout distance, I defined a bout of movement to be a phase of uninterrupted movement between two pauses or between a pause and a reversal. Thus each bout velocity represents the bout distance divided by the bout duration. Bouts in which the filament was moving at the start or end of the movie were ignored because their true duration could not be assessed. Statistical comparisons were performed using the Mann-Whitney test.

67

2.4.3 Imaging of EB1-YFP movement

To image EB1-YFP comets, cortical neurons were observed after 8-10 days in culture by epifluorescence microscopy using a Nikon TE2000 microscope with a 100x Plan Apo VC

1.4NA oil immersion objective. For these experiments, the exciting light from the mercury arc lamp was attenuated 4-fold using neutral density filters, and images were acquired with one second exposures at two second intervals for two minutes using a

CoolSNAP HQ cooled CCD camera and 2 × 2 pixel binning (Roper Scientific).

2.4.4 Analysis of EB1-YFP movement

Movement of EB1-YFP was analyzed by analyzing kymographs. Generating a kymograph means converting a time-lapse movie along a line of interest (along the axon) into a position versus time image (Welzel et al., 2011; Zhang et al., 2011; Zhou et al.,

2001). In my study, I always drew the line of interest from proximal to distal axon.

Particles moving along this line appeared as trajectories with their slope equaling to their velocity. Based on the way I drew the line, the trajectory of anterograde movement appeared as a descending slope from upper left to lower right and the trajectory of retrograde movement appeared as a descending slope from upper right to lower left. The 68 advantage of using kymographs for particle movement analysis is the data reduction

(Zhang et al., 2011) and graphical representation of particle movements in a single image

(Miller and Sheetz, 2004; Miller and Sheetz, 2006). In addition, kymographs increase the signal-to-noise ratio by averaging or taking the maximum of some pixels around the line of interest. For each time-lapse movie, the direction and velocity of each movement was acquired by analyzing the kymograph generated from the movie. Because EB1-YFP marks the growing end of microtubule (plus end), anterograde movement of EB1-YFP means that the plus end of microtubule points to the end of the axon.

2.5 Immunocytochemistry

Cortical neurons were transfected with GFP-NFM either with or without cMyc-N256S- kinesin-1A by electroporation. Cells were immunostained ten or eleven days after plating. For immunostaining, cells were first washed twice with PBS and then fixed with

4% paraformaldehyde. After fixation, cells were extracted with PBS containing 1% [v/v]

Triton X-100 and 0.3 M NaCl. After extraction, cells were stained first with a rabbit polyclonal specific for NFM (AB1987; Chemicon, MA, 1:200) and then subsequently

69 with a mouse monoclonal antibody specific for beta-tubulin (N357; Amersham, NJ,

1:400) and a goat polyclonal antibody specific for GFP (Goat anti-GFP; Rockland,

1:2000). The secondary antibodies were Alexa 488-donkey anti-goat (Invitrogen, 1:200),

Alexa 568-donkey anti-mouse (Invitrogen, 1:200), and Alexa 647-donkey anti-rabbit

(Invitrogen, 1:200). Actin was visualized by including Alexa-568 phalloidin (Invitrogen,

1:20) in the secondary antibody mixture. Coverslips were mounted by using 70 µl

ProLong Gold Antifade reagents (Invitrogen).

Images were acquired on a Nikon TE2000 microscope with a 40x Plan Apo 1.0 NA oil immersion objective and a CoolSNAP HQ cooled CCD camera. The epifluorescent illumination was attenuated 4-fold using neutral density filters, and images were acquired with 100 millisecond exposures. In case of cMyc-N256S-kinesin-1A, transfected cells were identified as cells expressing GFP-NFM. To quantify the fluorescence in the distal axon, I measured the fluorescence intensity in the most distal 100 μm of each axon, extending proximally from the base of the growth cone. Statistical comparisons were performed using the Mann-Whitney test.

70

2.6 Immunoprecipitation (IP)

2.6.1 Immunoprecipitation from mouse brain homogenate

50 µl Dynabeads® Protein G magnetic beads (Invitrogen) were coated with 10 µg

(typically 1 µg/µl) primary antibody (diluted in 200 µl PBS + 0.02% Tween-20), for 10 minutes at room temperature. The primary antibodies used in the present study were mouse monoclonal anti-NFM antibody (RMO270, Invitrogen), MAB1614 (clone H2,

Millipore, Billerica, MA) for kinesin-1 and mouse monoclonal anti-p150Glued antibody

(BD Transduction Laboratories, Sparks, MD). After coating with the antibodies, the beads were washed three times with 500 µl lysis buffer containing 20 mM Tris.HCl (pH

7.4), 1 mM EGTA, 1 mM EDTA, 0.15 M NaCl and 10 μg/ml protease inhibitor cocktails including bestatin (Sigma), E64 (Sigma) and leupeptin (Sigma).

An entire brain from a postnatal day 8-10 wild type ICR mouse (male or female), which was an offspring from the pregnant mouse obtained from Harlan Laboratories, was homogenized in 5 ml lysis buffer using POLYTRON PT2100 (Analytical Instruments,

Golden Valley, MN) at a speed of 20,000 rpm, and then centrifuged for 40 minutes at 71

3000 g at 4°C. After centrifugation, 200 µl supernatant was applied to the beads coated with primary antibodies. The mixture was incubated for an hour in the cold room (4°C) with constant rotating.

For conventional IP experiments, supernatant was collected for further analysis after separating the beads with a magnetic stand. The beads were washed three times with 200

µl lysis buffer per wash. Proteins were eluted from the beads using 250 µl 0.1M citric acid (pH 2-3), twice. This 500 µl pellet was concentrated to less than 30 µl using Amicon

Ultra 0.5 ml centrifugal filters (10K device, Millipore). Both supernatant and pellet were denatured at 100°C for 5 min in sample buffer containing 125 mM Tris pH 6.8, 2% sodium dodecyl sulfate (SDS; Sigma), 0.0025% Bromophenol Blue (Sigma), 7%

Glycerol (Sigma) and 5% β-mercaptoethanol (Sigma).

For sequential IP experiments, supernatant separated from the first IP was applied to new beads coated with the same antibody as used in the first IP. Conditions were tested to ensure that proteins that were capable to bind to the antibody were all precipitated in the first IP. In other words, the supernatant from the first IP was immuno-depleted by the

72 antibody. After the second IP, supernatant was collected by separating the beads with a magnetic stand. Binding proteins from both first and second IPs were separately eluted from the beads and denatured using the same method as described above for conventional

IP for further analysis.

2.6.2 Immunoprecipitation from HEK293T cells

HEK293T cells were transfected with DNA constructs of interest one day before IP experiments. On the day of IP, cells were first wash once with warm PBS, and then another time with cold PBS. Cells in each 60 mm dish were lysed in 1 ml lysis buffer containing 20 mM Tris.HCl (pH 7.4), 1 mM EGTA, 1 mM EDTA, 0.15 M NaCl, 1%

Nonidet P40 (Roche, Indianapolis, IN) and 10 μg/ml protease inhibitor cocktails including bestatin, E64 and leupeptin. Cells were scraped from the dish with a cell scraper and then homogenized manually using a 2-ml glass/teflon homogenizer. After homogenization, the lysate was centrifuged at 5000 g for 30 minutes and the supernatant was used in subsequent IP experiments.

73

In the IP experiments, the supernatant mentioned above was loaded to 50 μl beads coated with 10 μg antibody of interest. The mixture was incubated at room temperature for one hour with constant shaking. After incubation, beads were separated from the supernatant.

Binding proteins were de-natured for further analysis by adding 40 μl sample buffer and heating at 100°C for 5 min.

2.7 Western Blotting

SDS-PAGE was performed using 7.5% resolving gel, containing 7.5% acrylamide

(Sigma), 0.2% bisacrylamide (Sigma), 375 mM Tris (pH 8.8), 0.1% SDS, 0.1% ammonium persulfate (APS; Sigma), 0.1% tetramethylethylenediamine (TEMED;

Sigma), and 3.5% stacking gel, containing 3.5% acrylamide, 0.5% bisacrylamide, 125 mM Tris (pH 6.8), 0.1% SDS, 0.1% APS, 0.1% TEMED, 0.001% bromophenol blue, via

Hoefer gel running system (Hoefer, Holliston, MA). Running current for stacking gel was

0.02A and for resolving gel was 0.04A. Gels were transferred overnight to PVDF membranes (Immobilon-P Transfer Membrance, pore size 0.45 µm, Millipore) in the cold

74 room using Bio-Rad mini transfer cell (Bio-Rad, Hercules, CA). The voltage for transfer was 30V.

After transfer, membranes were first rinsed in methanol (Fisher Scientific, Pittsburg, PA)

and then washed three times with ddH2O before blocking in 5% [w/v] non-fat milk (non- fat dry milk (KrogerTM, Columbus, OH) dissolved in Tris Buffered Saline (TBS; Sigma)).

After blocking for 30 minutes, membranes were incubated with primary antibodies for an hour at room temperature. NFL was detected using a rabbit polyclonal antibody (1:5000) acquired from Dr. Virginia Lee (University of Pennsylvania, Philadelphia, PA)

(Trojanowski et al., 1989). NFM was detected using a mouse monoclonal antibody

RMO270 (Invitrogen) (1:5000), which binds NFM in a phospho-independent manner

(Yan et al., 2007). NFH was detected with AB1989 (Millipore) (1:1000), which binds

NFH in a phospho-independent manner. Kinesin-1A was detected using a rabbit polyclonal anti-kinesin-1A antibody (K0889, Sigma) (1:500). p150 was detected with anti-p150Glued antibody (1:1000). GFP-kinesin-5 was detected with a rabbit polyclonal anti-GFP antibody (Invitrogen) (1:1000). Cherry-vimentin was detected with a monoclonal anti-vimentin antibody (V6630, Sigma) (1:1000).

75

After primary antibody incubation, membranes were washed three times with TBST

(TBS plus 0.05% Tween-20 (Sigma)) and then three times with TBS. After blocking the membrane for another 15 minutes in 5% non-fat milk, secondary antibodies were applied to the membranes for an hour at room temperature.

For enhanced chemiluminescence (ECL) detection, secondary antibodies were either

HRP (horseradish peroxidase) conjugated goat anti-rabbit IgG or HRP conjugated goat anti-mouse IgG (1:10,000). After washing three times in TBST and three times in TBS, membranes were developed using ECL Plus™ Western Blotting Detection Reagents (GE

Healthcare, Piscataway, NJ) and X-ray films (Phenix Research Products, Candler, NC).

For KPL detection, secondary antibodies were either alkaline phosphatase conjugated goat anti-rabbit IgG or alkaline phosphatase conjugated goat anti-mouse IgG (Jackson

ImmunoResearch Laboratories, PA) (1:1000). After washing three times in TBST and three times in TBS, membranes were developed using BCIP/NBT Phosphatase Substrate

System (3-Component) kit (KPL, Gaithersburg, MD).

76

X-ray films and membranes were scanned using a Microtek Scan-maker i900 flatbed scanner (Microtek, Cerritos, CA). For quantification of the percentages of total proteins precipitated in the pellet, the intensities of the corresponding bands were measured using

MetaMorph software and corrected for background by background subtraction.

2.8 Immunostaining of native neurofilaments and motor proteins in vitro

Preparation of native neurofilaments from mouse or rat spinal cord

Native neurofilaments were prepared by Dr. Uchida in the Brown lab. The procedure is briefly summarized below. The starting materials were spinal cord tissues either from 5 mice or 1 rat. The ages of the mice were around 18-20 weeks. Spinal cords were chopped into 2 mm segments after being dissected out. Tissues were treated twice with 10 ml pH

6.2 hypotonic buffer, containing 10 mM sodium phosphate buffer (pH 6.2~6.8), 2 mM

EGTA, protease inhibitor cocktails (10 ug/ml leupeptin, 10 ug/ml E64, 10 ug/ml Bestatin,

1 ug/ml Aprotinin and 0.7ug/ml Pepstatin A) and PhosSTOP phosphatase inhibitor cocktails (Roche), on ice for 30 minutes per treatment. Tissues were then treated twice with 10 ml pH 6.8 hypotonic buffer on ice for 30 minutes per treatment. After hypotonic 77 treatments, tissues were rinsed with 5 ml RB buffer, containing 100 mM MES (pH 6.8)

(Sigma), 5 mM EGTA, 1 mM MgCl2 (Sigma) as well as protease and phosphatase inhibitor cocktails. After rinsing, equal volume of RB buffer to the volume of the tissues was added to the tissues and the tissues were manually homogenized using glass/teflon homogenizer.

After homogenization, tissues were briefly centrifuged at 5,000 g for 5 minutes at 4℃.

The supernatant was loaded to tubes containing 1.5 M sucrose in RB (1.5 ml) on top of

2.0 M sucrose in RB (100 μl) and centrifuged at 55,000 rpm (average 280,000 g, maxim

330,000 g) for 14 hours at 4℃ using SW 55 rotor (Beckman Coulter, Taunton, MA).

After centrifugation, native neurofilaments were recovered from the loose pellet in 2.0 M sucrose layer.

Preparation of DETA coated coverslips

1% Trimethoxysilylpropyldiethylenetriamine (DETA, Sigma) was hydrolyzed in 1 mM acetic acid (Fisher Scientific) for 5 minutes. Acid washed coverslips were immersed in this solution for 2 minutes and then rinsed in de-ionized water for 10 minutes. Finally, 78 they were dried for 15 minutes at 150 °C. Coverslips were good to use within one month.

The coverslip was attached to the bottom of tissue culture dish which had a hole in the middle using paraffin.

Immunostaining procedure

Neurofilaments purified from spinal cords were diluted to desired concentration in RB buffer and then applied to DETA coated coverslips in a total volume of 250 µl. After one minute, coverslips were rinsed twice with PBS and then blocked for 30 minutes with blocking solution containing 5% non-fat milk and 5% fish gelatin (Sigma) in PBS.

Primary antibodies used in this study are mouse monoclonal anti-NFM antibody

(RMO270, 1:400) and rabbit polyclonal anti-kinesin-1A antibody (K0889, 1:200).

Primary antibodies were applied for 45 minutes at 37℃. After washing three times with

PBS and blocking for another 15 minutes, secondary antibodies were applied to the coverslips for 45 minutes at 37℃. Secondary antibodies used in this study are Alexa-488 goat anti-mouse IgG and Quantum dot-605 goat anti-rabbit IgG. After washing three times with PBS, coverslips were immersed in 2 ml PBS for observation.

79

Chapter 3 The role of a hereditary spastic paraplegia mutation

in kinesin-1A/KIF5A in neurofilament transport

This chapter was reproduced from Wang and Brown 2010 paper, which was written by both authors.

3.1 Introduction

Neurofilaments are transported fast intermittently along microtubules propelled by fast motors. Several lines of evidence suggest that kinesin-1 may be the anterograde motor for neurofilaments. Shea and colleagues claimed that kinesin-1 interacted with neurofilament

M and H subunits in a phosphorylation-dependent manner (Jung et al., 2005; Yabe et al.,

2000). In addition, kinesin-1 has been shown to co-localize with non-filamentous neurofilament protein particles in squid axoplasm (Prahlad et al., 2000) and microinjection of kinesin-1 antibodies results in retention of peripherin in the cell body, and depletion of peripherin from the neurites in PC12 cells (Helfand et al., 2003) as well

80 as prevents the assembly of an extended vimentin network in spreading BHK-21 cells

(Prahlad et al., 1998). Although there was accumulating evidence linking kinesin-1 to intermediate filament transport, there was lack of evidence specifying which isoform of kinesin-1 was critical in this process.

A recent study revealed that kinesin-1s are homodimers of heavy chains and light chains

(DeBoer et al., 2008). In mammals, there are three conventional kinesin heavy chain genes and two conventional kinesin light chain genes, which give raise to six variant forms of conventional kinesins. Since these variants associated with biochemically different membrane-bounded organelles, it is possible that each variant transports a specific group of cargoes. Recently, two studies started to reveal the isoform of kinesin-1 responsible for anterograde neurofilament transport.

The first study came from Larry Goldstein’s lab using kinesin-1A knockout mice (Xia et al., 2003). The conventional knockouts died shortly after birth, so the authors created conditional knockouts by using a Cre-lox strategy, in which Cre recombinase expression was driven by the synapsin-I promoter, which turns on several weeks after birth. Seventy-

81 five percent of these mice exhibited seizures and died at around three weeks of age, but the twenty-five percent remaining lived for at least 3 months. These mice exhibited neurofilament accumulations in the cell body of sensory neurons accompanied by a striking reduction in sensory axon caliber. In contrast, the movement of several vesicle markers was not affected. Based on these observations, the authors hypothesized that kinesin-1A is a neurofilament motor. However, neurofilament transport was not studied directly in this study.

The first direct evidence suggesting that kinesin-1A is the principle motor for neurofilaments came from a study from the Brown lab (Uchida et al., 2009). In this study, the authors investigated neurofilament transport in cultured sympathetic neurons from wild type and kinesin-1A knockout mice using live-cell fluorescence imaging. The authors observed a 75% reduction in the frequency of neurofilament transport in both anterograde and retrograde directions in neurons cultured from kinesin-1A knockout mice, which could be rescued by expression of kinesin-1A and with decreasing efficacy by expression of kinesin-1B/C. In addition, expression of a headless kinesin-1A/C, which lacked the motor domain of kinesin-1, inhibited neurofilament transport in both directions

82 in cultured wild type neurons in a dominant-negative manner. These data suggest that kinesin-1A is the principle motor but not the exclusive motor for neurofilament transport at least in sympathetic neurons and it is necessary for both anterograde and retrograde neurofilament movement.

In this thesis, I have focused on investigating the effect of a kinesin-1A mutation, which causes hereditary spastic paraplegia (Please refer to Introduction 1.5 for more information about hereditary spastic paraplegia), on neurofilament transport. To date, at least 41 spastic paraplegia gene loci have been mapped (termed SPG1 through SPG 41) and 17 genes have been identified (Salinas et al., 2008). The inheritance can be autosomal dominant, autosomal recessive, or X-linked. One of the autosomal dominant forms,

SPG10, is caused by mutations in kinesin-1A. Of the 16 different SPG10 mutations that have been identified to date, 15 reside in the kinesin motor domain and 14 of these are missense mutations (Ebbing et al., 2008; Fichera et al., 2004; Miller and Sheetz, 2006;

Myers and Baas, 2007; Reid et al., 2002; Welzel et al., 2011; Zhang et al., 2011). A particular hot spot for these mutations is in the vicinity of the microtubule and nucleotide binding sites. In this thesis, I have focused on the N256S mutation, which results in the

83 substitution of a highly conserved asparagine residue in the switch II loop/helix motif of the microtubule binding site (Reid et al., 2002) (Figure 3.1). I investigated the transport of neurofilaments in cultured mouse cortical neurons over-expressing wild type or N256S kinesin-1A using live-cell imaging.

84

Figure 3.1 Sequence comparison of mouse and human kinesin-1A Sequence alignment of kinesin-1A in mice (Genbank Accession number NP_001034089) and humans (Genbank Accession number NP_004975). The motor domain spans residues 1 through 336 and is identical except for conserved substitutions at residues 41 and 218. The N256S mutation occurs at an invariant asparagine residue (highlighted in inverted contrast) located in the switch II loop/helix motif of the microtubule binding site.

3.2 Observation of neurofilament transport

To study the effect of N256S-kinesin-1A on neurofilament transport, I transfected mouse cortical neurons with GFP-tagged neurofilament protein M (GFP-NFM) either with or

85 without mutant or wild type mouse kinesin-1A by electroporation. The purpose of the wild type kinesin-1A was to control for any possible effects due to over-expression of the motor. Mouse cortical neurons exhibit gaps in the axonal neurofilament array similar to those observed in neurons from superior cervical ganglia (Perez-Olle et al., 2005; Uchida et al., 2009; Wang et al., 2000), but the gaps in the cortical neurons are longer and more numerous, making these cells particularly suitable for studies of neurofilament movement

(Figure 3.2). The GFP-NFM fusion protein incorporates throughout all the neurofilaments in these cells, permitting all neurofilaments to be detected.

86

Figure 3.2 Neurofilament moving through a gap An axon of a cortical neuron expressing GFP-NFM was visualized by epifluorescence microscopy. Cortical axons exhibit discontinuities in the axonal neurofilament array, called gaps. Filaments that move into the gaps can be tracked by time-lapse imaging to analyze the kinetics of movement. Proximal is left and distal is right. Scale bar = 10 μm.

87

3.3 Neurofilament transport frequency and flux

To track the movement of the GFP-tagged neurofilaments, I observed gaps by epifluorescence microscopy and acquired time-lapse movies using one second exposures at four second time intervals. Each movie was exactly 15 minutes in length. Ninety six percent of the moving structures were filamentous in shape, ranging from 1.3 μm to 44.5

μm in length and diffraction limited in width. The average length was 8.6 μm, which is comparable to what the Brown lab observed in previous studies on neurons from mouse superior cervical ganglia (Uchida et al., 2009; Uchida and Brown, 2004). I defined transport frequency as the number of filaments that moved at least 50 pixels (6.55 μm) per 15-minute movie. N256S-kinesin-1A reduced the average neurofilament transport frequency significantly, from 4.5 to 1.4 filaments/hour in the anterograde direction (p <

0.001) and from 3.2 to 2.0 filaments/hour in the retrograde direction (p = 0.047) (Figure

3.3). Expression of wild type kinesin-1A reduced the average neurofilament transport frequency from 4.5 to 3.7 filaments/hour anterogradely and from 3.2 to 3.1 filaments/hour retrogradely, but these effects were not statistically significant (p = 0.26

88 and p = 0.92, respectively). Thus N256S-kinesin-1A impaired neurofilament transport in both anterograde and retrograde directions in these axons.

To analyze the motility defect in more detail, I tracked the movement of each filament through successive frames of the time-lapse movies. I defined anterograde and retrograde transport flux as the total distance moved in the corresponding direction by all the filaments in each 15-minute movie. Since each movie contained a single axon, I expressed the fluxes in units of μm/axon/hour. N256S-kinesin-1A reduced the average neurofilament transport flux from 149 to 46 μm/axon/hour in the anterograde direction (p

< 0.001) and from 116 to 90 μm/axon/hour in the retrograde direction (p = 0.007; Figure

3.3). Expression of wild type kinesin-1A reduced the transport flux from 149 to 109

μm/axon/hour anterogradely and from 116 to 105 μm/axon/hour retrogradely, but these reductions were not statistically significant (p = 0.13 and p = 0.69, respectively). Thus, in addition to decreasing the frequency of neurofilament movement, N256S-kinesin-1A also decreased the total extent of neurofilament movement.

89

Figure 3.3 Effect of N256S-kinesin-1A on the frequency and flux of neurofilament movement Analysis of neurofilament movement through gaps in axons of cortical neurons expressing GFP-NFM alone (A & D; 51 movies), GFP-NFM plus N256S-kinesin-1A (B & E; 40 movies), or GFP-NFM plus exogenous wild type kinesin-1A (C & F; 50 movies). (A-C) Histograms of the frequencies of neurofilament movement, expressed as number of neurofilaments per hour. The filaments were classified as anterograde or retrograde based on their preferred direction of movement and an anterograde and retrograde frequency was calculated for each movie. (D-F) Histograms of the neurofilament fluxes, expressed as total distance moved by all the filaments per axon per hour in either the anterograde or retrograde direction. An anterograde and retrograde flux was calculated for each movie.

90

3.4 Neurofilament transport bout velocity, distance and duration

To determine whether N256S-kinesin-1A also affected the velocity or persistence of neurofilament movement, I measured the velocity, distance and duration of each bout of movement for each moving filament. I defined a bout as a period of uninterrupted movement between two pauses or between a pause and a reversal. N256S-kinesin-1A decreased the average bout velocity from 0.27 μm/s to 0.24 μm/s in the anterograde direction, but this was not statistically significant (p = 0.053; Figure 3.4). In contrast,

N256S-kinesin-1A increased the average retrograde bout velocity from 0.32 to 0.40 μm/s, and this was statistically significant (p < 0.001). N256S-kinesin-1A also increased the average retrograde bout distance from 3.7 μm to 5.7 μm (p < 0.001) and the average retrograde bout duration from 11 seconds to 15 seconds (p < 0.001), but without any significant effect on average anterograde bout distance (3.2 μm, p = 0.36) or average anterograde bout duration (19 seconds, p = 0.98). Expression of wild type kinesin-1A had no significant effect on average retrograde bout velocity (0.30 μm/s, p = 0.76), average retrograde bout distance (3.4 μm, p = 0.57), or average retrograde bout duration (14 seconds, p = 0.32). Expression of wild type kinesin-1A did decrease the average

91 anterograde bout velocity from 0.27 μm/s to 0.23 μm/s (p = 0.006) and average anterograde bout distance from 3.4 μm to 3.0 μm (p = 0.043), but without any significant effect on average anterograde bout duration (15 seconds, p = 0.31).

92

Figure 3.4 Effect of N256S-kinesin-1A on the velocity, distance and duration of bouts of neurofilament movement Analysis of the velocity, distance and duration of bouts of neurofilament movement in axons of cortical neurons expressing GFP-NFM alone (A, D & G; 246 anterograde bouts, 212 retrograde bouts), GFP-NFM plus N256S-kinesin-1A (B, E & H; 84 anterograde bouts, 91 retrograde bouts), or GFP-NFM plus exogenous wild type kinesin-1A (C, F & I; 254 anterograde bouts, 223 retrograde bouts). The anterograde and retrograde bouts are represented by the dark grey and light grey bars, respectively. The y-axis represents the percentage of bouts in the corresponding direction. Note: to avoid compressing the scale on the x-axis, the right-most bin represents an expanded bin size of 30-50 in graphs D-F, and 100-500 in graphs G-I. 93

Thus, N256S-kinesin-1A reduced the anterograde flux by decreasing anterograde frequency without affecting anterograde velocity. N256S-kinesin-1A also reduced the retrograde flux in spite of an increase in retrograde velocity because the increase in retrograde velocity was not sufficient to compensate for the decrease in retrograde frequency (summarized in Figure 3.5).

94

Figure 3.5 Summary of kinetic data Bar graphs summarizing the kinetics of neurofilament movement in axons of cortical neurons expressing GFP-NFM alone, GFP-NFM plus N256S-kinesin-1A, or GFP-NFM plus exogenous wild type kinesin-1A. Graphs A-D summarize the data in Figure 3.3 and graphs E-J summarize the data in Figure 3.4. The asterisks denote the p values comparing the GFP-NFM plus N256S-kinesin-1A and the GFP-NFM plus exogenous wild type kinesin-1A to GFP-NFM alone (Mann-Whitney test: *** = p < 0.005; ** = p < 0.01; * = p < 0.05). 95

3.5 Comparison of neurofilament content of distal axons

It has been shown previously that neurofilaments are delivered to distal axons by anterograde movement and retrieved by retrograde movement (Uchida and Brown, 2004).

Thus it is possible that the decrease in retrograde flux described above could be explained by a depletion of neurofilaments from distal axons as a secondary consequence of the disruption of anterograde movement. To test this hypothesis, I transfected cultured cortical neurons with GFP-NFM, either with or without N256S-kinesin-1A, and then fixed and processed the cells for immunofluorescence microscopy after 10-11 days in culture using antibodies specific for GFP and neurofilament protein M (NFM). Since axonal neurofilament distribution is discontinuous in these neurons, I also stained for tubulin and actin, which are present along the entire length of each axon. Cells expressing

N256S-kinesin-1A were identified by their GFP expression. To quantify neurofilament content, I measured the fluorescence intensity of NFM in distal axons, extending 100 μm proximally from the base of the growth cone. There was no statistically significant difference in the neurofilament content of axons expressing N256S-kinesin-1A compared to axons expressing no exogenous kinesin-1A (Figure 3.6). Thus, disruption of

96 neurofilament transport by N256S-kinesin-1A does not appear to deplete distal axons of neurofilaments, and therefore the reduction in retrograde neurofilament flux cannot be a secondary consequence of the disruption of anterograde neurofilament movement. The absence of a reduction in neurofilament content in distal axons expressing N256S- kinesin-1A is probably due to the impairment of both anterograde and retrograde neurofilament movement, which would be expected to impair both delivery and departure of neurofilaments from these axonal regions.

97

Figure 3.6 N256S-kinesin-1A does not deplete neurofilaments from distal axons Comparison of the neurofilament content of distal axons from cortical neurons expressing GFP-NFM alone (control) or GFP-NFM plus N256S-kinesin-1A. (A) Immunofluorescence microscopy for neurofilament protein (top panels) and for tubulin and actin (bottom panels). Scale bar = 20 μm. (B) Quantification of neurofilament content in distal axons. There was no significant difference in the neurofilament content of axons expressing GFP-NFM plus N256S-kinesin-1A (49 axons) compared to control axons expressing GFP-NFM alone (49 axons; p = 0.73, Mann-Whitney test).

98

3.6 Microtubule polarity in axons of cultured mouse cortical neurons

Microtubules in axons are widely accepted to be orientated exclusively with their plus- ends distal, and kinesin-1 motors are known to move cargoes exclusively toward the plus- ends of these polymers (Baas et al., 1988; Baas et al., 1987; Burton and Paige, 1981;

Heidemann et al., 1981). Thus the disruption of retrograde neurofilament movement by

N256S-kinesin-1A suggests that this mutant also disrupts minus-end directed neurofilament movement, which is thought to be mediated by dynein (He et al., 2005;

Shah et al., 2000; Wagner et al., 2004). However, another possibility, albeit unlikely, is that microtubule polarity in cultured mouse cortical axons is not entirely or predominantly plus-end distal. To test this hypothesis, I transfected cultured cortical neurons with the microtubule plus-end tracking protein EB1 tagged with yellow fluorescent protein (EB1-YFP) and imaged the movement of EB1-YFP comets. Using kymograph analysis, I tracked 192 comets in 23 axons. 190 comets moved anterogradely and 2 comets moved retrogradely (Figure 3.7). The average comet velocity was 0.13

μm/s, which is consistent with published estimates of the rate of microtubule growth in cells (Shubeita et al., 2008). These data confirm that microtubules in these axons are

99 indeed almost exclusively plus-end distal (especially considering that the plus-end proximal microtubules could represent microtubules that looped back on themselves).

Thus the disruption of retrograde neurofilament transport by N256S-kinesin-1A appears to be due to disruption of minus-end directed movement by this mutant plus-end directed motor.

100

Figure 3.7 Microtubules in cortical neuron axons are plus-end distal Cortical neurons were transfected with EB1-YFP, which is a protein that binds to the growing plus ends of microtubules. Axons were imaged for 2 minutes at 2 second intervals in order to determine the direction of microtubule growth. (A) A typical kymograph showing a number of EB1-YFP comets moving anterogradely. The horizontal dimension represents distance and the vertical dimension represents time. (B) Histogram of EB1-YFP comet velocities (192 comets from 23 different neurons). Note that almost all the comets move anterogradely, confirming the plus-end distal orientation of the microtubules in these axons.

101

3.7 Summary

In summary, the present data reveal that expression of N256S-kinesin-1A in cultured mouse cortical neurons impaired neurofilament transport in both anterograde and retrograde directions. This was due primarily to a decrease in the frequency, not the velocity of movement. A limitation of my experimental approach is that transient transfection of the mutant motor does not permit quantification or regulation of the expression level relative to the endogenous wild type motor. Therefore, to control for possible effects of over-expression, I also characterized neurofilament movement in neurons transfected with wild type kinesin-1A motor. Simply over-expressing wild type kinesin-1A had a small effect on anterograde bout velocity and anterograde bout distance, but no effect on average frequency or flux in either the anterograde or retrograde direction. Thus the effects of the N256S-kinesin-1A on neurofilament transport were due to the N256S mutation and were not an artifact of over-expression.

The impairment of both anterograde and retrograde neurofilament movement in these experiments is notable because kinesin-1A is an anterograde motor in axons, but this

102 result is consistent with recent evidence from the Brown laboratory showing that the anterograde and retrograde neurofilament motors are interdependent (Uchida et al.,

2009). In that study, the authors found that both anterograde and retrograde neurofilament movement were impaired in neurons from kinesin-1A knockout mice, and that expression of wild type kinesin-1A rescued the movement in both directions. In addition, expression of a headless dominant negative kinesin-1A construct in wild type neurons impaired both anterograde and retrograde neurofilament movement, and disruption of dynein function by using RNA interference, dominant negative approaches, or a function-blocking antibody also inhibited both anterograde and retrograde neurofilament movement. In this thesis, I confirmed that the reduction in retrograde movement was not due to depletion of neurofilament content at the distal region of axons. In addition, I confirmed that the polarity of microtubules in cortical axons is plus-end distal, so the retrograde defect of neurofilament movement could not be driven by anterograde motor. Taken together, these data suggest that there is functional coupling between kinesin-1A and dynein motors in the bidirectional transport of neurofilaments along microtubules in axons.

103

In the present study, I found that expression of N256S-kinesin-1A reduced the frequency and flux of neurofilament movement in anterograde and retrograde directions and increased bout velocity, distance and duration in the retrograde direction. It is unclear how to interpret these data mechanistically. The decrease in both anterograde and retrograde frequency and flux suggests a coordination mechanism, in which the opposing motors interact so that only motors of one directionality are bound or active at one time, but the increase in retrograde bout velocity, distance and duration suggests a tug-of-war, in which the direction of movement is the result of a dynamic competition between opposing motors that are bound and active at the same time. Thus it is likely that the mechanism is more complex, perhaps combining features of both models (See Chapter 5 for discussion).

104

Chapter 4 Interaction of neurofilaments with kinesin-1A and

dynein/dynactin

4.1 Introduction

Today, it is widely accepted that neurofilaments move along axons driven by fast motors.

However, compared to the vast majority of live-cell imaging studies of neurofilament transport under manipulations of different motor components, direct evidence demonstrating interactions between motors and neurofilaments is rare. There is slightly more literature on the interaction between the neurofilament and its retrograde motor dynein/dynactin compared to the interaction between the neurofilament and its anterograde motor kinesin. Direct evidence showing the interaction between neurofilaments and dynein/dynactin came mainly from two studies.

In the first study, Shah et al. (Shah et al., 2000) purified neurofilaments from bovine spinal cords and probed the fractions with antibodies to components of dynein/dynactin

105 complex. The authors observed strong reactivity of neurofilaments with several components of dynein/dynactin complex, including dynein intermediate chain, dynein heavy chain, p150glued, p50/dynamitin and capZ. The interaction was not disrupted by 1%

Triton X-100, indicating that it did not involve membranous structures. In addition, they observed heterogeneous co-localization of dynein intermediate chain immunoreactivity with neurofilaments using immuno-electron microscopy with some regions showing dense labeling with dynein intermediate chain and other regions showing sparse labeling.

In this study, the authors failed to identify conventional kinesin in purified neurofilaments. Using an anti-peptide antibody developed against the microtubule binding region of Xenopus conventional kinesin heavy chain, they identified a number of putative kinesin related proteins. However, the identities of these proteins were not clear.

In a more recent study from Wanger et al. (Wagner et al., 2004), the authors visualized decorations of cytoplasmic dynein/dynactin complex along the length of neurofilaments using atomic force microscopy. In addition, they identified a direct interaction between dynein intermediate chain and NFM using yeast two-hybrid and affinity chromatography assays.

106

Compared to the data indicating the interaction between neurofilaments and dynein/dynactin, evidence showing an interaction between neurofilaments and kinesin is very rare and mostly unclear. Most of the data claiming an interaction between neurofilaments and kinesin came from studies in Thomas Shea’s lab. In an earlier study from this lab, the authors found that NFL, NFM and NFH co-purified and co-precipitated with motors from motor preparation, in which motors were co-sedimented with microtubules by adding AMP-PNP, which forces the binding of motors to microtubules and then the motors were released by adding ATP to the mixture (Yabe et al., 1999). In this paper, the blotting results were very messy. Several bands were identified by pan- kinesin antibody in the motor preparation. The conventional kinesin antibody used in this experiment identified another 175 kDa band, which was also identified by the pan- kinesin antibody, in addition to conventional kinesin heavy chain. Therefore, it was not clear whether neurofilament subunits were associated with conventional kinesin or the other 175 kDa protein in their experiments.

In two follow-up papers from the same lab, the authors claimed that the interaction between kinesin and neurofilaments depended on the phosphorylation status of

107 neurofilaments (Jung et al., 2005; Yabe et al., 2000). They found that kinesin preferentially interacted with non-phosphorylated neurofilament subunits. Again, in these studies, kinesin was identified by a pan-kinesin antibody and the quality of the data was not very good. Until now, there is no evidence clearly demonstrating an interaction between neurofilaments and kinesin-1, not to mention kinesin-1A. One of the reasons is that previously there was no well characterized antibody against each kinesin-1 isoform.

In this thesis, I characterized the interaction between kinesin-1A and neurofilaments using immunoprecipitation assays with well characterized kinesin-1A antibodies under several well designed controls. To confirm the interaction between neurofilaments and dynein/dynactin, I investigated whether the p150 subunit, a component of dynactin, interacted with neurofilaments since dynein usually functions together with dynactin.

Identifying the interaction between neurofilaments and motors is very challenging due to the following reasons. Firstly, at any given time, 97% of neurofilaments are not moving

(Trivedi et al., 2007). It is possible that these stationary neurofilaments are not bound to motors at all. Secondly, recent studies suggest that a few motors may be sufficient to move a cargo (Shubeita et al., 2008). Finally, motors have multiple cargoes besides

108 neurofilaments. Because of these reasons, I hypothesized that the percentage of motor proteins coming down with neurofilaments would be very small and vice versa.

4.2 Characterization of kinesin-1 antibodies

In mammals, there are three kinesin-1 (KIF5) heavy chain genes, kinesin-1A, 1B and 1C.

Kinesin-1A and kinesin-1C are expressed in neurons, whereas kinesin-1B is expressed ubiquitously. I first characterized several kinesin-1 antibodies I had using HEK293T cells, which do not express endogenous kinesin-1A or kinesin-1C. To characterize the antibodies, HEK cells were transfected with kinesin-1A, kinesin-1B or kinesin-1C DNA construct by lipofection. One day after transfection, cell lysates were analyzed by

Western blot. H2 antibody recognized kinesin-1A and kinesin-1C, with higher affinity to kinesin-1A. Although kinesin-1B was present in abundance in the lysate from cells

transfected with kinesin-1B construct, it was not detected by H2 antibody (Figure 4.1).

The kinesin-1B specific antibody UIC-81 was a generous gift from Dr. Scott Brady at

University of Illinois at Chicago (DeBoer et al., 2008).

109

Figure 4.1 Characterization of H2 antibody by Western blot Kinesin-1A, 1B or 1C was separately expressed in HEK cells. Cell lysates were analyzed by Western blot (marked as kin-1A, kin-1B and kin-1C on top). Upper panel: H2 antibody detected kinesin-1A (higher affinity) and kinesin-1C, but not kinesin-1B, although kinesin-1B was present in abundance (middle panel). The weaker band at higher molecular weight was a non-specific binding of UIC-81. GAPDH was used as a loading control (lower panel).

When I blotted mouse brain homogenate using H2 antibody, it recognized a doublet corresponding to kinesin-1A (upper band, ~117 KDa) and kinesin-1C (lower band, ~109

KDa). The lower band could be a combination of kinesin-1B (~110 KDa) and kinesin-1C

110 although very unlikely based on my observations in HEK cells. Due to the similarity in molecular weight between kinesin-1B (109.5 KDa) and kinesin-1C (109.3 KDa), they could not be distinguished in SDS-PAGE. In contrast, when I performed

immunoprecipitation using H2 antibody from mouse brain homogenate, only kinesin-1A

proteins were precipitated in the pellet. This may be due to the higher affinity of H2

antibody to kinesin-1A (Figure 4.2). Therefore, H2 antibody was used as a kinesin-1A

specific antibody in terms of immunoprecipitation in my study. In addition to H2 antibody, I characterized a commercially available anti-kinesin-1A antibody (K0889), which recognizes amino acid residues 1007-1027 of mouse kinesin-1A. This antibody specifically recognizes kinesin-1A, without cross reactivity with kinesin-1B or kinesin-

1C (Figure 4.2). Therefore, it was used in Western blot to probe kinesin-1A.

111

Figure 4.2 Characterization of H2 antibody in immunoprecipitation Using HEK cells expressing either headless kinesin-1A, full length kinesin-1A, kinesin- 1B or kinesin-1C, I have shown that K0889 antibody interacts specifically with full length and headless kinesin-1A without any cross reactivity with kinesin-1B or kinesin-

1C (left). Immunoprecipitation experiments were performed using H2 antibody coated beads from mouse brain homogenate. As shown in the upper right panel, H2 only precipitated kinesin-1A in the pellet (P: pellet), although both kinesin-1A and kinesin-1C were present in the homogenate (S: supernatant). The identity of the band in the pellet was further confirmed by K0889 antibody as shown in the lower right panel, in which only kinesin-1A was specifically detected in both pellet and supernatant fractions.

4.3 Interaction between neurofilaments and motor proteins revealed by conventional immunoprecipitation (IP) from mouse brain homogenate

Neurofilaments were immunoprecipitated from postnatal day 8–10 mouse brain homogenate using phospho-independent NFM antibody (RMO 270) coated magnetic beads. Pellets (P) were analyzed by Western blot. Besides NFM, NFL and NFH were also

112 precipitated in the pellet (Figure 4.3) due to the reason that neurofilaments are obligate heteropolymers in vivo (Lee et al., 1993). Therefore, neurofilament subunits are co- assembled to form filaments in vivo. I quantified the percentage of each neurofilament subunit proteins in the pellet. The average percentage of NFL in the pellet was 100% of total NFL proteins in the homogenate (n = 3). The average percentage of NFM in the pellet was 94% of total NFM proteins (n = 9). The average percentage of NFH in the pellet was 46% of total NFH proteins (n = 3) (Figure 4.4). The reason why I only precipitated about 50% of total NFH proteins is probably because during early development, a significant amount of NFH proteins are in soluble form (Sanchez et al.,

2000), which could not be precipitated with NFM antibody. NFM antibody can precipitate NFL and NFH proteins only when they co-assemble with NFM to form neurofilaments. In this study, mouse IgG coated beads were used as negative controls.

For controls, neurofilament subunit proteins were only detected in the supernatant (S).

There was no detectable neurofilament protein in the pellet. These results demonstrated that neurofilaments were efficiently immunoprecipitated in the pellet using the NFM antibody.

113

Next, I asked whether motor proteins were co-immunoprecipitated with neurofilaments.

Since I expected that the percentage of motor proteins coming down with neurofilaments would be very small. To make the detection easier, I concentrated the protein in the pellet by 1-3 fold and loaded much less supernatant than pellet. On average, there was about

0.08% kinesin-1A (n = 5) and 0.53% p150 (n = 6) co-immunoprecipitated with neurofilaments (Figure 4.3 and 4.4). For the control, there was no detectable kinesin-1A or p150 in the pellet.

114

Figure 4.3 IP with NFM antibody or IgG coated beads NFM antibody precipitated neurofilament triplet proteins as well as a small amount of kinesin-1A and p150 in the pellet (P). Loading for neurofilament proteins was equivalent in P and supernatant (S). Loading for kinesin-1A was 200 times more in P than in S, for p150 was 80 times more in P than in S. For IgG coated beads (negative controls), all proteins remained in S.

115

IP NFM 120 1.2

100 1

80 0.8

60 0.6

40 0.4

% of % protein total pellet in 20 0.2

0 0 Blot NFL NFM NFH kin-1A p150

Figure 4.4 Quantification of proteins in the pellet Using NFM antibody, on average, I precipitated 100% NFL (n = 3), 94% NFM (n = 9) and 46% NFH (n = 3) in the pellet. Together with neurofilaments, 0.08% kinesin-1A (n = 5) and 0.53% p150 (n = 6) were co-precipitated in the pellet.

To verify these interactions, I performed reciprocal IP with antibodies against motor proteins. A proportion of kinesin-1A proteins were precipitated with H2 antibody.

Together with kinesin-1A, a small percentage of NFM proteins were co-precipitated

(Figure 4.5). The average percentage of kinesin-1A precipitated by H2 antibody in the pellet was 55% (n = 5). On average, there was about 0.13% NFM (n = 7) coming down

116 with kinesin-1A in the pellet (Figure 4.5). For controls, NFM and kinesin-1A proteins remained in the supernatant.

Figure 4.5 IP with H2 antibody or IgG coated beads Using H2 antibody, I precipitated an average of 55% kinesin-1A (kin-1A) in the pellet (n = 5). Together with kinesin-1A, there was about 0.13% NFM co-precipitated in the pellet (n = 7). Loading for kinesin-1A was equivalent in P and S. Loading for NFM was 200 times more in P than in S. For IgG coated beads (negative control), all proteins remained in S.

Besides H2 antibody, I also performed reciprocal IP with p150 antibody. A proportion of p150 proteins were precipitated with p150 antibody. Together with p150, a small percentage of NFM proteins were co-precipitated (Figure 4.6). The average percentage of

117 p150 precipitated by p150 antibody was 36% of total p150 proteins in the homogenate (n

= 3). On average, there was about 0.94% NFM (n = 3) coming down with p150 in the pellet (Figure 4.6).

Figure 4.6 IP with p150 antibody or IgG coated beads Using p150 antibody, I precipitated an average of 36% p150 in the pellet (n = 3). Together with p150, there was about 0.94% NFM co-precipitated in the pellet (n = 3). Loading for p150 was 4 times more in P than in S. Loading for NFM was 200 times more in P than in S. For IgG coated beads (negative control), all proteins remained in S.

118

4.4 Interaction between neurofilament and motor proteins revealed by sequential IP from mouse brain homogenate

Since the amount of motor proteins coming down with neurofilaments were very small and vice versa, there is a possibility that this small proportion of motor proteins may come down in the pellet due to non-specific interaction with neurofilament antibody and vice versa. To test this possibility, I designed a sequential IP approach. In this experiment, I first applied mouse brain homogenate to beads coated with antibodies I was interested in, for example NFM antibody. I performed condition tests to ensure that all neurofilaments that were able to interact with the antibody were precipitated in the first

IP in pellet 1 (P1). Supernatant from the first IP was added to new beads coated with the same antibody, in this case NFM antibody. If the interaction between neurofilaments and motors is specific, when there was no neurofilament in P2, there should not be any motor in P2. This also applies to sequential IP with motor antibodies.

For sequential IP using NFM antibody, almost all NFM proteins were precipitated in P1.

There was no NFM protein left in P2 or S2 (supernatant after the second IP). Together

119 with NFM, there was a small amount of kinesin-1A and p150 coming down in P1. When there was no NFM protein in P2, there was no kinesin-1A and p150 in P2 either, although both kinesin-1A and p150 proteins were present in abundance in S2 (Figure 4.7).

Figure 4.7 Sequential IP with NFM antibody Almost all NFM proteins were precipitated in pellet 1 (P1). There was no detectable NFM in pellet 2 (P2) and supernatant 2 (S2). A small amount of p150 (lower left) and kinesin-1A (lower right) were co-precipitated with NFM in P1. When there was no NFM in P2, there was no p150 or kinesin-1A in P2, although both p150 and kinesin-1A proteins were abundant in S2. Loadings for NFM were 1-2 times more in P1 and P2 than in S2. Loading for p150 was 80 times more in P1 and P2 than in S2 and loading for kinesin-1A was 40 times more in P1 and P2 than in S2.

120

I also performed reciprocal sequential IP using antibodies against motor proteins. In case of IP using antibodies against motors, a proportion of motor proteins were precipitated in

P1. Small amount of NFM proteins were co-precipitated with motor proteins in P1.

Although a proportion of motor proteins were still left in S2, I could not precipitate more motor proteins in P2. One possible explanation could be that motor proteins might have different states in native homogenate. For example, kinesin-1A tail domain could interact with the head domain to inhibit the activity of the motor. It is possible that some confirmations of the motor might prevent the accessibility of the antibodies.

Nevertheless, when there was no precipitated in P2, there was no NFM protein in P2 (Figure 4.8). These results demonstrated that the interaction between neurofilament and motors is specific.

121

Figure 4.8 Sequential IP with p150 or H2 antibody Left: sequential IP with p150 antibody. A proportion of p150 proteins were precipitated in P1. Although there were still a proportion of p150 proteins left in S2, I could not precipitate more p150 protein in P2. Together with p150, there were a small proportion of NFM proteins co-precipitated. Loading for p150 was 4 times more in P1 and P2 than in S2. Loading for NFM was 200 times more in P1 and P2 than in S2. When there was no p150 in P2, there was no NFM in P2, although there were abundant NFM proteins in S2. Right: sequential IP using H2 antibody. A proportion of kinesin-1A proteins were precipitated in P1. Although there were still a proportion of kinesin-1A proteins left in S2, I could not precipitate more kinesin-1A in P2. Together with kinesin-1A, there were a small proportion of NFM proteins co-precipitated. Loading for kinesin-1A was equivalent in P1, P2 and S2. Loading for NFM was 180 times more in P1 and P2 than in S2. When there was no kinesin-1A in P2, there was no NFM in P2.

4.5 Interaction between neurofilament and motors confirmed by IP using neurofilaments purified from mouse/rat spinal cords

To confirm the interaction between neurofilament and motor proteins, I also performed

IP using native neurofilaments purified from mouse or rat spinal cords as inputs.

122

Neurofilaments were highly enriched in this preparation, which made the detection of the interaction much easier compared with mouse brain homogenate. Neurofilaments purified from mouse or rat spinal cords were added to beads coated with either NFM antibody or control IgG. Both kinesin-1A and p150 were co-purified and co-precipitated with neurofilaments from the preparation (Figure 4.9), which confirms the phenomena I observed using mouse brain homogenate. Moreover, when 0.75 M NaCl was added to the preparation before immunoprecipitation, the interaction between neurofilaments and motors was reduced (Figure 4.9), which is in agreement with previous findings from

Shah et al. that 1 M KCl reduced the interaction between neurofilaments and dynein intermediate chain (Shah et al., 2000).

It was possible that the decrease of interaction between neurofilaments and motor proteins under high salt condition was caused by disruption of neurofilament structures, for example disruption of neurofilament polymers, instead of disruption of the interaction between neurofilaments and motors. To test this possibility, I attached neurofilaments under either high salt or normal condition to glass coverslips coated with DETA and immunostained neurofilaments with NFM antibody. My results showed that there was no

123 distinguishable morphological difference between untreated and high salt treated neurofilaments (Figure 4.10), which demonstrated that high salt treatment disrupted the interaction between neurofilaments and motors without affecting the morphology of neurofilaments.

124

Figure 4.9 Comparison of 0.75 M NaCl treated and untreated groups Neurofilaments were treated with 0.75 M NaCl for an hour and then precipitated with NFM antibody. Compared to untreated group, there were less kinesin-1A and p150 co- precipitated with neurofilaments, although the amount of neurofilaments precipitated in both group were comparable. In the p150 blot, the lower band is an alternative splicing form of DCTN1, the gene that encodes p150Glued, called p135 with a molecular weight of 135 kDa (Dixit et al., 2008). In contrast, IgG coated beads could not precipitate either neurofilaments or motors.

125

Figure 4.10 Immunostaining of neurofilaments in NaCl treated and untreated groups Neurofilaments purified from mouse spinal cord were treated with 0.75M NaCl for an hour before absorbing to coverslips. Neurofilaments were stained using an antibody against NFM (RMO 270) and visualized with Alexa 488 conjugated secondary antibodies. Left: Neurofilaments without NaCl treatment. Right: Neurofilaments with 0.75 M NaCl treatment. (Scale bar = 5 µm)

4.6 Interaction between kinesin-1A and each neurofilament subunit

To further confirm the interaction between kinesin-1A and neurofilaments and test for potential subunit preference in this interaction, I investigated the interaction between kinesin-1A and each neurofilament subunit (NFL, NFM and NFH) in HEK 293T cells.

126

There are several advantages of using this cell line. The transfection efficiency of HEK cells is very high. The expression level of exogenous proteins in these cells is usually very high, which makes the detection much easier. In addition, HEK cells lack kinesin-

1A and neurofilaments, which provides us a clean background to investigate the interaction between kinesin-1A and each neurofilament subunit.

In this study, cells were transfected with kinesin-1A plus one of the three neurofilament subunit constructs. Since HEK cells have their own vimentin intermediate filament networks, the exogenous neurofilament proteins may co-assemble with endogenous vimentin to form filaments. Alternatively, the exogenous neurofilament proteins could also exist in soluble forms. Immunoprecipitations were performed with H2 antibody. The results indicate that kinesin-1A interacts with each neurofilament subunit, preferentially with NFM (Figure 4.11).

To control for any possible false positives caused by nonspecific interactions between motors and intermediate filaments, I used vimentin and kinesin-5 as another level of controls besides IgG control. Vimentin has a similar structure as neurofilaments with a

127 coiled coil domain and kinesin-5 shares a similar structure with kinesin-1A with a coiled coil domain. For control experiments, I co-expressed cherry-vimentin and kinesin-1 in

HEK cells. I performed IP with H2 antibody and then blotted for vimentin. I also co- expressed EGFP-kinesin-5 and NFM in HEK cells. I performed IP with NFM antibody and then blotted for GFP. In either case, vimentin or kinesin-5 was not precipitated in the pellet (Figure 4.11), suggesting that the interactions between kinesin-1A and each neurofilament subunit were specific.

128

Figure 4.11 Interactions between kinesin-1A and neurofilament subunits All three neurofilament subunits co-precipitated with kinesin-1A (P). In case of control IgG, all proteins remained in the supernatant (S). For NFL blot, I used alkaline phosphatase based detection method instead of ECL. Vimentin did not co-precipitate with kinesin-1A. For vimentin blot, the upper band represents cherry-vimentin and the lower band represents endogenous vimentin. EGFP-kinesin-5 did not co-precipitate with NFM neither.

129

4.7 Summary

In the present study, I identified for the first time an interaction between neurofilaments and kinesin-1A and confirmed the interaction between neurofilaments and dynein/dynactin. The percentage of motor proteins interacting with neurofilaments at any given time was very small and vice versa. This was not surprising to us, although it made the detection of the interaction challenging. By concentrating the proteins in the pellet and loading much less supernatant than pellet, I was able to clearly identify the interaction. I performed reciprocal IPs by precipitating motor proteins, kinesin-1A and p150, and confirmed the interaction between neurofilaments and motor proteins.

Since the percentage of motor proteins co-precipitated with neurofilaments was very small and vice versa, I performed sequential IPs to rule out the possibility that these small amount of motor proteins may interact non-specifically with neurofilament antibody and vice versa. My sequential IP results revealed that when there was no neurofilament protein in pellet 2, there was no motor protein in pellet 2, although there were abundant

130 motor proteins in supernatant 2. My reciprocal sequential IPs yielded similar results.

Therefore, the interactions between neurofilaments and motor proteins were specific.

In addition to studies using mouse brain homogenate, I also performed IP experiments using native neurofilaments purified from mouse or rat spinal cords. In this preparation, neurofilaments were highly enriched, which made my detection of the interaction much easier. Besides confirming the interaction between neurofilaments and motor proteins using this purified neurofilament prep, I also found that when the prep was treated with

0.75 M NaCl before IP, the interactions between neurofilaments and motor proteins were reduced.

Compared to the interaction between dynein/dynactin and neurofilaments, which was much better characterized, the interaction between kinesin-1A and neurofilaments was identified for the first time. Therefore, I further characterized this interaction. Firstly, I tested whether this interaction has any preference to any of the three neurofilament subunits. By co-expressing kinesin-1A heavy chain and each neurofilament subunit construct in HEK cells, I performed IPs to investigate the interaction between kinesin-1A

131 and each neurofilament subunit. My results showed that kinesin-1A interacted with each neurofilament subunit possibly through a direct interaction. In other words, the interaction may not depend on kinesin light chain. We should note that HEK cells have endogenous kinesin light chains which may be involved in this interaction. Therefore, further evidence is needed to make a conclusion of whether the interaction is direct or indirect. (See discussion in chapter 5). In addition, my results suggest that kinesin-1A preferentially binds to NFM compared with NFL and NFH.

132

Chapter 5 Discussion

Section 5.1.1, 5.1.2 and 5.2.1 are excerpted verbatim from Wang and Brown 2010 paper written by both authors.

5.1 Kinesin-1A and neurofilament transport

5.1.1 N256S-kinesin-1A mutation and neurofilament movement

In this thesis, I found that N256S-kinesin-1A mutant decreased both frequency and flux of neurofilament transport in both anterograde and retrograde directions. Studies in vitro suggest that the N256S-kinesin-1A mutant is a defective motor and that it may act as a dominant-negative disruptor of kinesin-1A transport (Ebbing et al., 2008). The motor domains of kinesins are highly conserved across species. Mutation of the homologous amino acid residue to a lysine in the motor domain of the yeast kinesin-14 motor kar3

(N650K-kar3) or in the motor domain of the fungal kinesin-14 motor ncd (N600K-ncd), prevents microtubule-stimulated activation of the motor ATPase by uncoupling

133 nucleotide and microtubule binding (Song and Endow, 1998). Both N650K-kar3 and

N600K-ncd are capable of binding and hydrolyzing ATP, but in contrast to wild type kinesin, they bind tightly to microtubules in both the ATP-bound and ADP-bound states.

In microtubule gliding assays in vitro, N600K-ncd motors bind microtubules but do not translocate them. In yeast cells, N650K-kar3 exhibits a dominant negative effect over wild-type kar3 (Hoyt et al., 1993). In vitro, N256S-kinesin-1A is unable to generate single-motor processive motion in microtubule gliding and bead motility assays, resulting in decreased gliding and transport velocities (Ebbing et al., 2008). However, in contrast to the N to K mutations in N650K-kar3 and N600K-ncd, the N256S mutation in kinesin-

1A did not bind microtubules in rigor. When N256S-kinesin-1A was mixed with wild type kinesin-1A in microtubule gliding and particle motility assays in vitro, the mutant kinesin appeared to exert a dominant inhibitory effect on the transport velocity. Given these observations, however, it is surprising that the disruption of neurofilament transport by N256S-kinesin-1A in my study was due primarily to a decrease in the frequency.

There was a slight decrease in average velocity (from 0.27 to 0.24 μm/s), but this was not statistically significant. Moreover, this effect was probably an artifact of over-expression, since I observed a similar decrease in cells expressing wild type kinesin-1A. Thus the

134 behavior of the motor in motility assays in vitro cannot necessarily predict its effect on cargo transport in vivo.

5.1.2 Models of bi-directional cargo movement

The mechanism by which microtubule motors of opposing directionality interact to regulate bidirectional cargo transport is not yet understood. Two favored models are the tug-of-war model and the coordination model (regulated switching) (Welte, 2004)

(Figure 5.1). In the tug-of-war model, the direction of movement is the result of a dynamic competition between opposing motors that are bound and active at the same time. In the coordination model, the opposing motors interact so that only motors of one directionality are bound or active at one time. In their simplest form, these two models have quite different predictions: in the tug-of-war model, impairment of motors of one directionality should increase the velocity and frequency of movement in the opposite direction, whereas in the coordination model it should not. In fact, according to the coordination model, manipulations or mutations that disrupt the coordination could also cause the motors to interfere with each other, leading to impairment of movement in both directions (Gross, 2004). Many labs have reported reciprocal inhibition of both directions 135 of movement after disruption of motors of one directionality, which is consistent with the coordination model (see (Uchida et al., 2009) for citations to these studies).

136

Figure 5.1 Models for bidirectional cargo transport Many cargoes in eukaryotic cells have both kinesin and dynein motors on them and move in a bidirectional manner. Two models have been proposed to explain bidirectional movement: the regulated switching model (the coordination model) and the tug-of-war model. In the regulated switching model, the activity of opposite motors is regulated so that at any time only motors of one directionality are active. In the tug-of-war model, the opposing motors generate force against each other so that the direction of movement is determined by the winner (Verhey et al., 2011).

137

In the present study, I found that expression of N256S-kinesin-1A reduced the frequency and flux of neurofilament movement in anterograde and retrograde directions and increased bout velocity, distance and duration in the retrograde direction. It is unclear how to interpret these data mechanistically. The decrease in both anterograde and retrograde frequency and flux suggests a coordination mechanism, but the increase in retrograde bout velocity, distance and duration suggests a tug-of-war. Thus it is likely that the mechanism is more complex, perhaps combining features of both models. For example, Lipowsky and colleagues have shown that tug-of-war models that account for the load-dependence of the interaction between motors and their tracks can generate bouts of persistent anterograde and retrograde movement, depending on the relative numbers of bound motors (Muller et al., 2008; Muller et al., 2010). To test such hypotheses in the case of neurofilaments it will be necessary to record neurofilament movement with much higher spatial and temporal resolution than I have done in the present study, and also to measure the forces acting on the moving filaments, which is currently not possible because neurofilaments are too small to be optically trapped.

138

5.1.3 Interaction between kinesin-1A and neurofilaments

In the present study, I have identified an interaction between neurofilaments and kinesin-

1A using immunoprecipitation assays. The interaction demonstrated in mouse brain homogenate was under native protein state with native protein concentration. The amount of motors interacting with neurofilaments appeared to be very low and vice versa. The calculation of the percentages of total proteins in the pellet was relatively rough because of the non-linearity of the ECL detection method. Moreover, I could not actually quantify molar ratio from Westerns.

The low percentages of interactions were expected due to the primary reason stated below. Neurofilaments are transported in slow component a (SCa) with velocity ranging about 0.1-1 mm/day. The slow rate is a result of rapid movements interrupted by prolonged pauses. It has been estimated that neurofilaments spend 97% of their time pausing when they move along the axon. It is likely that these stationary neurofilaments are not interacting with motor proteins. As a result, under native protein state and concentration, such as in mouse brain homogenate, the interaction between neurofilament and motor proteins was not at an equal molar ratio. For example, we assume there are 139 about 30 polypeptides per unit length of the neurofilament (~60 nm) and the average length of neurofilaments is about 8 μm. Therefore, there are about 4000 polypeptides per neurofilament. If there are about 3% of neurofilaments moving at any given time and each filament was moved by one motor, then the molar ratio of the neurofilament to the motor will be 4000 divided by 3%, which equals to about 130,000 to 1. This was the primary reason that made the detection of the interaction very hard. By concentrating the pellets, I was able to clearly detect this interaction. I am aware that it is also possible that stationary neurofilaments could associate with inactive motor proteins and the movement of neurofilament could be a result of regulation of the activity of motors instead of regulating motor/neurofilament interaction, although my data favor the latter mechanism.

In the present study, I used H2 antibody in my IP experiments. Based on my characterizations, H2 antibody detected both kinesin-1A and kinesin-1C in Western blot.

However, in terms of IP, it only precipitated kinesin-1A, although all three isoforms of kinesin-1 existed in mouse brain homogenate. Previously, H2 antibody had been used as a pan-specific kinesin-1 antibody, which supposed to recognize all three isoforms of kinesin-1 (Pfister et al., 1989). However, based on my observations in HEK cells, H2

140 antibody primarily recognized kinesin-1A and kinesin-1C, likely with higher affinity to kinesin-1A. It was possible although very unlikely that the expression level of kinesin-1A and kinesin-1C might be different since kinesin-1A and kinesin-1C were subcloned in the same vector and the expression was driven by the same promoter. H2 antibody did not detect kinesin-1B even though it was over-expressed. Previous studies from the Brady lab acquired similar results. The author found that H2, which recognizes amino acid sequence 400-600 of mouse kinesin-1s, binds primarily to kinesin-1A and kinesin-1C, but not much to kinesin-1B (DeBoer et al., 2008). In terms of IP from mouse brain homogenate, although all three isoforms of kinesin-1 existed, H2 only precipitated kinesin-1A in the pellet, which was probably due to its higher affinity to kinesin-1A, therefore my IP data with H2 antibody were specific to kinesin-1A isoform. DeBoer et al. also found that H2 precipatated kinesin-1s with different affinities. I tried to perform IP using the other kinesin-1A specific antibody (K0889), but it did not bind native kinesin-

1A with high affinity. It is possible that this antibody may not interact with native kinesin-1A very well, although it is very good for detection of kinesin-1A in Western blot.

141

Based on previous studies from the Brown lab (Uchida et al., 2009), it seems that kinesin-

1A could not be the only motor for neurofilaments, although it seems to be the principle motor for neurofilaments at least in cultured mouse sympathetic neurons. In the present interaction study, I did not investigate the potential interactions between other isoforms of kinesin-1 and neurofilaments. It is possible that neurofilaments might also interact with kinesin-1B and kinesin-1C. However, based on previous live-cell imaging studies from the Brown lab, I speculated that the interactions between neurofilaments and kinesin-1B/C might be a lot weaker if they do exist. Since the interaction between neurofilaments and kinesin-1A was already not easy to detect, my current methodology may not be suitable to detect much weaker interaction. In the present study, I did try to blot kinesin-1C from the pellet precipitated using NFM antibody. I was not able to detect any kinesin-1C signal in the pellet, although this did not necessarily prove that kinesin-

1C did not bind to neurofilaments. I am aware that my current assays may not be powerful enough to detect very weak signals. Nevertheless, my data clearly reveal an interaction between kinesin-1A and neurofilament.

142

In addition to the data acquired using mouse brain homogenate, I also confirmed the interactions between neurofilaments and kinesin-1A and dynein/dynactin using purified neurofilaments from mouse and rat spinal cords. This preparation was highly enriched with neurofilaments. As a result, I did not have to concentrate the IP pellet. My results clearly revealed interactions between neurofilaments and kinesin-1A and dynein/dynactin in this preparation. Moreover, the interactions were greatly reduced by manipulating salt concentration of the preparation, which added another level of control for specificity.

One downside of my current IP assay is that it does not provide us any information about the directness of the interaction. While there was some evidence showing that the interaction between neurofilaments and dynein/dynactin was mediated through direct interaction between dynein intermediate chain and NFM, there is no such evidence for kinesin-1A. Therefore, I further characterized the interaction between neurofilaments and kinesin-1A in HEK cells. My results suggest that kinesin-1A preferentially interacts with

NFL and NFM, to less extend with NFH. I am aware that this difference could be caused by different expression levels of neurofilament subunit proteins in HEK cells. This difference could also be caused by different post-translational modifications of each

143 neurofilament subunit. For example, it is possible that NFH was heavily phosphorylated, resulting in preventing the interaction between NFH and kinesin-1A, although I did not access the phosphorylation status of each neurofilament subunit in my study. Due to the difference between HEK cell system and physiological condition, results I acquired from

HEK cells do not necessarily predict the phenomena in vivo. Nevertheless, the data I got from HEK cells does confirm the interaction between kinesin-1A and neurofilaments in another level.

Unlike previous studies on the interaction between neurofilaments and dynein/dynactin, in which the authors used multiple approaches to prove and visualize the interaction, my current results on the interaction between neurofilaments and kinesin-1A were merely based on immunoprecipitation approach, which is a great approach to begin with since it reveals protein interactions under physiological conditions. However, in terms of IP, identification of the interaction depends greatly on antibodies and the stableness of the interaction. Moreover, it cannot provide us more details on the interaction. Therefore, I would like to demonstrate and visualize the interaction in the future using other approaches, such as immunostaining, live-cell imaging approaches (See 5.3 for details).

144

5.2 Kinesin-1A and hereditary spastic paraplegia

5.2.1 Neurofilament and HSP

Neurofilaments accumulate abnormally and excessively in many neurodegenerative diseases, including amyotrophic lateral sclerosis, giant axonal neuropathy, and Charcot

Marie Tooth disease (Al-Chalabi and Miller, 2003; Perrot and Eyer, 2009). Several studies have suggested that these accumulations may arise due to perturbations in axonal transport (Chevalier-Larsen and Holzbaur, 2006; Collard et al., 1995; De Vos et al.,

2008). In support of this idea, slowing of axonal transport has been reported in mouse models of SOD-mediated amyotrophic lateral sclerosis (Zhang et al., 1997) and is an early event in the progression of this disease (Williamson and Cleveland, 1999).

Disruption of axonal transport by over-expression of dynamitin in neurons also results in accumulations of neurofilaments (LaMonte et al., 2002), and mutations in dynein subunits are one cause of motor neuron disease in humans (Hafezparast et al., 2003; Puls et al., 2003). Defects in axonal transport have also been reported in mouse models of spastin-mediated hereditary spastic paraplegia (SPG4) (Kasher et al., 2009; Tarrade et al.,

2006), and swellings containing vesicular and cytoskeletal proteins, including 145 neurofilament proteins, have been reported in humans with this disease (Kasher et al.,

2009; Tarrade et al., 2006).

While it is clear that the axonal transport of neurofilaments is impaired in many neurodegenerative diseases, the role of neurofilament accumulations in the disease progression has been controversial. Over-expression of the human high molecular weight neurofilament protein (NFH) in mice causes a slowing of neurofilament transport, accumulations of axonal neurofilaments, and motor neuron degeneration (Collard et al.,

1995; Cote et al., 1993). However, the significance of these studies is unclear because over-expression of mouse NFH has no pathological effects (Marszalek et al., 1996). One approach to test the role of neurofilaments in mouse models of neurodegenerative diseases has been to cross the mice with neurofilament L knockout mice, which lack neurofilament polymers. Using this approach, it has been shown that the absence of neurofilaments in neurons slows the progression of disease dramatically in mouse models of superoxide dismutase (SOD)-mediated ALS (Lobsiger et al., 2005; Williamson et al.,

1998). On the other hand, similar experiments using a transgenic mouse expressing an

NFH β-galactosidase fusion protein, which aggregates and sequesters neurofilaments in

146 neuronal cell bodies, showed no significant alteration of disease pathology in mouse models of dystonia musculorum and SOD-mediated ALS, though there was some prolongation of neuronal survival and some delay of axon loss (Eyer et al., 1998).

Apparently contradictory results on the role of neurofilaments in disease have also been obtained in experiments on the accumulation of neurofilaments in response to neurotoxins such as acrylamide and hexanedione. These agents impair or accelerate axonal transport of neurofilaments and other cargoes and lead to focal accumulations and depletions of axonal neurofilaments. Studies with transgenic mice expressing NFH β- galactosidase fusion protein suggest that neurofilaments are not essential for the toxicity associated with the administration of these substances (Stone et al., 2001), whereas studies on the Quiverer (Quv) quail, which lack NFL, suggest that they are (Hirai et al.,

1999). Perhaps the conflicting nature of these reports may be due to differences between the animal models used. For example, the NFH β-galactosidase transgenic mice have perikaryal neurofilament accumulations whereas the NFL knockout mice and the

Quiverer quail do not. Whatever the explanation, however, it does seem clear that the

147 accumulation of neurofilaments can be an exacerbating factor in at least some circumstances.

Neurofilaments are unlikely to be the sole cargo for kinesin-1A in neurons so it is possible that deficiencies in the movement of other cargoes may contribute to the disease progression in SPG10 (See the section below for discussion). Moreover, while neurofilament accumulations have been described in patients with SPG4 (see above), there have been no ultrastructural studies on nerves of patients with SPG10. Thus it is presently unclear whether neurofilament accumulations are a feature of this disease. In the present study, I did not observe local neurofilament accumulations in axons of neurons expressing N256S-kinesin-1A. However, it is unclear to what extent such observations in short term cultures can predict the long-term effects of SPG10 mutaions on neurofilament organization in vivo. For example, it is quite possible that subtle changes in neurofilament organization or distribution in short-term cultures of neonatal cultured neurons might be magnified over longer time scales in mature neurons in vivo.

Either way, the fact that kinesin-1A appears to be a neurofilament motor and that N256S- kinesin-1A disrupts the bidirectional transport of neurofilaments in cultured neurons

148 suggest that patients with the SPG10 form of hereditary spastic paraplegia may well have neurofilament transport abnormalities which may contribute to the disease progression, and this warrants further investigation.

5.2.2 Other cargoes for kinesin-1A

In case of SPG10, the mutations are in kinesin-1A, which is one isoform of conventional kinesin. Although I have shown that neurofilament movement was disrupted in both directions when I over-expressed a SPG10 kinesin-1A mutation, I am aware that kinesin-

1A may transport other cargoes besides neurofilament. It is very unlikely that neurofilament is the only cargo for kinesin-1A. Till now, there is very few known potential cargoes and interactors of kinesin-1A, including HAP-1 (huntingtin associated protein-1) (Twelvetrees et al., 2010), DISC-1 (disrupted in schizophrenia protein-1) and the NUDEL/LIS1/14-3-3ε complex (Taya et al., 2007), Grb2 (growth factor receptor bound protein-2) (Shinoda et al., 2007), and β-dystrobrevin (Ceccarini et al., 2005;

Macioce et al., 2003). It is very possible that defects in transport of one or several of these cargoes might be involved in the pathogenesis of SPG10.

149

5.3 Future directions

5.3.1 Characterization and verification of the interaction between neurofilaments and kinesin-1A

Visualize the interactions between neurofilaments and motors by doing immunostaining

I have developed an approach to absorb neurofilaments onto glass coverslips (See

Chapter 2 for method). These coverslips were treated with DETA, which made them highly positively charged. Since neurofilaments are highly phosphorylated in vivo, they are negatively charged. As a result, DETA coated coverslips absorb neurofilaments very well. I tested several blocking conditions and found that blocking solution containing 5% non-fat milk plus 5% fish gelatin provided the lowest background. I used neurofilaments purified from mouse or rat spinal cords in this study because I have already shown that there were motors binding to these neurofilaments by doing immunoprecipitation. I used

Alexa-488 conjugated secondary antibody to visualize neurofilaments and Quantum dot-

605 conjugated secondary antibody to label kinesin-1A. I chose Quantum dot conjugated secondary antibody based on the assumption that a few motors may be sufficient to move 150 a filament. Since Quantum dot is much brighter and more photo-stable than Alexa dye, it may allow me to identify a few protein molecules. I expected that the percentage of neurofilaments having motors on them may be very low based on my IP results. A typical picture I acquired is shown in Figure 5.2. Usually, I observed a small percentage of neurofilaments in the field of view with motors on them. However, compared to the number of quantum dots in the background, it is hard to say whether the interaction I observed is specific or not at this time point. Because of the nature of this interaction

(low percentage), it is very difficult to discriminate specific from non-specific interaction even with a lower background. It will be very helpful if I can increase the percentage of the interaction, for example by adding exogenous kinesin-1A to neurofilaments. I am planning to express kinesin-1A in bacteria and then purify the proteins and add them to purified neurofilaments. I hope that in this way I may be able to increase the change of kinesin-1A and neurofilaments interaction which will make my detection easier. Besides visualize the interaction between neurofilaments and motors, I could also investigate whether the opposing motors, kinesin-1A and dynein/dynactin, are co-localized on neurofilaments, which may provide us a mechanistic explanation for the functional interdependence of the motors at molecular level.

151

Figure 5.2 Immunostaining of purified neurofilaments with the neurofilament and kinesin-1A antibodies Neurofilaments purified from mouse spinal cord were absorbed on a DETA coated coverslip. Neurofilaments were stained with an antibody against NFM (RMO 270) and visualized by Alexa-488 conjugated secondary antibody (upper left). Kinesin-1A was stained with anti-kinesin-1A antibody (K0889) and visualized by Quantum dot 605 conjugate secondary antibody (upper right). The overlay was shown in the lower panel. (Scale bar = 5 µm)

152

Visualize the interactions between neurofilament and motors by live-cell imaging

In the first step, I would like to observe movement of kinesin-1A. To achieve this goal, I have constructed a green fluorescent protein (GFP) tagged kinesin-1A plasmid. However, when it was expressed in cultured mouse cortical neurons, it was not bright enough to let me visualize the movement of the motor. I am working on making a triple GFP tagged kinesin-1A. I hope that when it is expressed in neurons, it will be bright enough for observation of movement. Ultimately, I would like to observed movement of kinesin-1A

(GFP tagged) and neurofilaments (Cherry tagged) at the same time by doing dual imaging. I would like to see the location of kinesin-1A on neurofilaments. I hypothesize that kinesin-1A binds to the ends of neurofilaments when they move based on fluorescent intensity analysis of moving neurofilaments. I am also interested in seeing how many kinesin-1As are usually needed to move a filament.

Mapping the binding domains of kinesin-1A and neurofilament

In addition to demonstrating an interaction between kinesin-1A and neurofilaments, I would like to further characterize this interaction. For example, is this interaction mediated by kinesin light chain? My studies in HEK cells transfected with kinesin-1A 153 heavy chain and one of the neurofilament subunits suggest that the interaction may not need kinesin light chain, although I should be aware that HEK cells have endogenous kinesin light chain proteins which may be involved in this interaction. If the interaction between kinesin-1A heavy chain and neurofilaments is direct, which domain of kinesin-1 heavy chain is involved in this interaction? I have some preliminary data showing that headless-kinesin-1A, which lacks the motor domain of kinesin-1A, interacts with neurofilaments. Moreover, this interaction between neurofilaments and headless-kinesin-

1A is stronger than the interaction between neurofilaments and full length kinesin-1A

(Figure 5.3). Based on this observation, the motor domain of kinesin-1A may not be involved in neurofilament binding. I speculate that the tail domain of kinesin-1A is critical for the interaction with neurofilaments. Since the tail can interact with the head

(motor) domain to auto-inhibit the activity of the motor, removing the motor domain may release the tail for neurofilament interaction, which explains why the interaction between headless-kinesin-1A and neurofilaments is stronger than full length kinesin-1A. I am working on testing the interaction between neurofilaments and kinesin-1A-tail domain as well as between neurofilaments and kinesin-1A-tailless construct. The results of these

154 experiments may answer the question whether the tail domain of kinesin-1A plays a critical role in its interaction with neurofilaments.

155

References

Ackerley, S., A.J. Grierson, J. Brownlees, P. Thornhill, B.H. Anderton, P.N. Leigh, C.E. Shaw, and C.C. Miller. 2000. Glutamate slows axonal transport of neurofilaments in transfected neurons. J Cell Biol. 150:165-176. Ackerley, S., P. Thornhill, A.J. Grierson, J. Brownlees, B.H. Anderton, P.N. Leigh, C.E. Shaw, and C.C. Miller. 2003. Neurofilament heavy chain side arm phosphorylation regulates axonal transport of neurofilaments. J Cell Biol. 161:489-495. Aizawa, H., Y. Sekine, R. Takemura, Z. Zhang, M. Nangaku, and N. Hirokawa. 1992. Kinesin family in murine central nervous system. J Cell Biol. 119:1287-1296. Al-Chalabi, A., and C.C. Miller. 2003. Neurofilaments and neurological disease. Bioessays. 25:346-355. Alami, N.H., P. Jung, and A. Brown. 2009. Myosin Va increases the efficiency of neurofilament transport by decreasing the duration of long-term pauses. J Neurosci. 29:6625-6634. Angelides, K.J., K.E. Smith, and M. Takeda. 1989. Assembly and exchange of intermediate filament proteins of neurons: neurofilaments are dynamic structures. J Cell Biol. 108:1495-1506. Averback, P. 1981. Unusual particles in motor neuron disease. Arch Pathol Lab Med. 105:490-493. Baas, P.W., J.S. Deitch, M.M. Black, and G.A. Banker. 1988. Polarity orientation of microtubules in hippocampal neurons: uniformity in the axon and nonuniformity in the dendrite. Proc Natl Acad Sci U S A. 85:8335-8339. Baas, P.W., L.A. White, and S.R. Heidemann. 1987. Microtubule polarity reversal accompanies regrowth of amputated neurites. Proc Natl Acad Sci U S A. 84:5272- 5276. Bajaj, N.P., and C.C. Miller. 1997. Phosphorylation of neurofilament heavy-chain side- arm fragments by cyclin-dependent kinase-5 and glycogen synthase kinase-3alpha in transfected cells. J Neurochem. 69:737-743.

156

Barry, D.M., S. Millecamps, J.P. Julien, and M.L. Garcia. 2007. New movements in neurofilament transport, turnover and disease. Exp Cell Res. 313:2110-2120. Beaulieu, J.M., J. Robertson, and J.P. Julien. 1999. Interactions between peripherin and neurofilaments in cultured cells: disruption of peripherin assembly by the NF-M and NF-H subunits. Biochem Cell Biol. 77:41-45. Block, S.M., L.S. Goldstein, and B.J. Schnapp. 1990. Bead movement by single kinesin molecules studied with optical tweezers. Nature. 348:348-352. Boltshauser, E., A. Bischoff, and W. Isler. 1977. Giant axonal neuropathy. Report of a case with normal hair. J Neurol Sci. 31:269-278. Bomont, P., L. Cavalier, F. Blondeau, C. Ben Hamida, S. Belal, M. Tazir, E. Demir, H. Topaloglu, R. Korinthenberg, B. Tuysuz, P. Landrieu, F. Hentati, and M. Koenig. 2000. The gene encoding gigaxonin, a new member of the cytoskeletal BTB/kelch repeat family, is mutated in giant axonal neuropathy. Nat Genet. 26:370-374. Borchelt, D.R., P.C. Wong, M.W. Becher, C.A. Pardo, M.K. Lee, Z.S. Xu, G. Thinakaran, N.A. Jenkins, N.G. Copeland, S.S. Sisodia, D.W. Cleveland, D.L. Price, and P.N. Hoffman. 1998. Axonal transport of mutant superoxide dismutase 1 and focal axonal abnormalities in the proximal axons of transgenic mice. Neurobiol Dis. 5:27-35. Brady, S.T. 1995. A kinesin medley: biochemical and functional heterogeneity. Trends Cell Biol. 5:159-164. Bridgman, P.C. 1999. Myosin Va movements in normal and dilute-lethal axons provide support for a dual filament motor complex. J Cell Biol. 146:1045-1060. Brownlees, J., S. Ackerley, A.J. Grierson, N.J. Jacobsen, K. Shea, B.H. Anderton, P.N. Leigh, C.E. Shaw, and C.C. Miller. 2002. Charcot-Marie-Tooth disease neurofilament mutations disrupt neurofilament assembly and axonal transport. Hum Mol Genet. 11:2837-2844. Bruce Alberts, A.J., Julian Lewis, Martin Raff, Keith Roberts, and Peter Walter. 2002. Molecular Biology of the Cell. Garland Science, New York. Burgess, S.A., M.L. Walker, H. Sakakibara, P.J. Knight, and K. Oiwa. 2003. Dynein structure and power stroke. Nature. 421:715-718. Burkhardt, J.K., C.J. Echeverri, T. Nilsson, and R.B. Vallee. 1997. Overexpression of the dynamitin (p50) subunit of the dynactin complex disrupts dynein-dependent maintenance of membrane organelle distribution. J Cell Biol. 139:469-484.

157

Burton, P.R., and J.L. Paige. 1981. Polarity of axoplasmic microtubules in the olfactory nerve of the frog. Proc Natl Acad Sci U S A. 78:3269-3273. Cai, D., A.D. Hoppe, J.A. Swanson, and K.J. Verhey. 2007. Kinesin-1 structural organization and conformational changes revealed by FRET stoichiometry in live cells. J Cell Biol. 176:51-63. Carden, M.J., J.Q. Trojanowski, W.W. Schlaepfer, and V.M. Lee. 1987. Two-stage expression of neurofilament polypeptides during rat neurogenesis with early establishment of adult phosphorylation patterns. J Neurosci. 7:3489-3504. Carpenter, S. 1968. Proximal axonal enlargement in motor neuron disease. Neurology. 18:841-851. Ceccarini, M., P. Torreri, D.G. Lombardi, G. Macchia, P. Macioce, and T.C. Petrucci. 2005. Molecular basis of dystrobrevin interaction with kinesin heavy chain: structural determinants of their binding. J Mol Biol. 354:872-882. Chevalier-Larsen, E., and E.L. Holzbaur. 2006. Axonal transport and neurodegenerative disease. Biochim Biophys Acta. 1762:1094-1108. Ching, G.Y., and R.K. Liem. 1993. Assembly of type IV neuronal intermediate filaments in nonneuronal cells in the absence of preexisting cytoplasmic intermediate filaments. J Cell Biol. 122:1323-1335. Ching, G.Y., and R.K. Liem. 1999. Analysis of the roles of the head domains of type IV rat neuronal intermediate filament proteins in filament assembly using domain- swapped chimeric proteins. J Cell Sci. 112 ( Pt 13):2233-2240. Colakoglu, G., and A. Brown. 2009. Intermediate filaments exchange subunits along their length and elongate by end-to-end annealing. J Cell Biol. 185:769-777. Collard, J.F., F. Cote, and J.P. Julien. 1995. Defective axonal transport in a transgenic mouse model of amyotrophic lateral sclerosis. Nature. 375:61-64. Cote, F., J.F. Collard, and J.P. Julien. 1993. Progressive neuronopathy in transgenic mice expressing the human neurofilament heavy gene: a mouse model of amyotrophic lateral sclerosis. Cell. 73:35-46. Couillard-Despres, S., Q. Zhu, P.C. Wong, D.L. Price, D.W. Cleveland, and J.P. Julien. 1998. Protective effect of neurofilament heavy gene overexpression in motor neuron disease induced by mutant superoxide dismutase. Proc Natl Acad Sci U S A. 95:9626-9630.

158

Cuenca, N., E. Fernandez, J. de Juan, J. Carreres, and C. Iniguez. 1987. Postnatal development of microtubules and neurofilaments in the rat optic nerve: a quantitative study. J Comp Neurol. 263:613-617. De Vos, K.J., A.J. Grierson, S. Ackerley, and C.C. Miller. 2008. Role of axonal transport in neurodegenerative diseases. Annu Rev Neurosci. 31:151-173. DeBoer, S.R., Y. You, A. Szodorai, A. Kaminska, G. Pigino, E. Nwabuisi, B. Wang, T. Estrada-Hernandez, S. Kins, S.T. Brady, and G. Morfini. 2008. Conventional kinesin holoenzymes are composed of heavy and light chain homodimers. . 47:4535-4543. Delisle, M.B., and S. Carpenter. 1984. Neurofibrillary axonal swellings and amyotrophic lateral sclerosis. J Neurol Sci. 63:241-250. Dietrich, K.A., C.V. Sindelar, P.D. Brewer, K.H. Downing, C.R. Cremo, and S.E. Rice. 2008. The kinesin-1 motor protein is regulated by a direct interaction of its head and tail. Proc Natl Acad Sci U S A. 105:8938-8943. Ding, J., E. Allen, W. Wang, A. Valle, C. Wu, T. Nardine, B. Cui, J. Yi, A. Taylor, N.L. Jeon, S. Chu, Y. So, H. Vogel, R. Tolwani, W. Mobley, and Y. Yang. 2006. Gene targeting of GAN in mouse causes a toxic accumulation of microtubule-associated protein 8 and impaired retrograde axonal transport. Hum Mol Genet. 15:1451- 1463. Ding, J., J.J. Liu, A.S. Kowal, T. Nardine, P. Bhattacharya, A. Lee, and Y. Yang. 2002. Microtubule-associated protein 1B: a neuronal binding partner for gigaxonin. J Cell Biol. 158:427-433. Dixit, R., J.R. Levy, M. Tokito, L.A. Ligon, and E.L. Holzbaur. 2008. Regulation of dynactin through the differential expression of p150Glued isoforms. J Biol Chem. 283:33611-33619. Donaghy, M., R.H. King, P.K. Thomas, and J.M. Workman. 1988. Abnormalities of the axonal cytoskeleton in giant axonal neuropathy. J Neurocytol. 17:197-208. Dong, D.L., Z.S. Xu, M.R. Chevrier, R.J. Cotter, D.W. Cleveland, and G.W. Hart. 1993. Glycosylation of mammalian neurofilaments. Localization of multiple O-linked N-acetylglucosamine moieties on neurofilament polypeptides L and M. J Biol Chem. 268:16679-16687. Dong, D.L., Z.S. Xu, G.W. Hart, and D.W. Cleveland. 1996. Cytoplasmic O-GlcNAc modification of the head domain and the KSP repeat motif of the neurofilament protein neurofilament-H. J Biol Chem. 271:20845-20852. 159

Ebbing, B., K. Mann, A. Starosta, J. Jaud, L. Schols, R. Schule, and G. Woehlke. 2008. Effect of spastic paraplegia mutations in KIF5A kinesin on transport activity. Hum Mol Genet. 17:1245-1252. Elder, G.A., V.L. Friedrich, Jr., P. Bosco, C. Kang, A. Gourov, P.H. Tu, V.M. Lee, and R.A. Lazzarini. 1998a. Absence of the mid-sized neurofilament subunit decreases axonal calibers, levels of light neurofilament (NF-L), and neurofilament content. J Cell Biol. 141:727-739. Elder, G.A., V.L. Friedrich, Jr., C. Kang, P. Bosco, A. Gourov, P.H. Tu, B. Zhang, V.M. Lee, and R.A. Lazzarini. 1998b. Requirement of heavy neurofilament subunit in the development of axons with large calibers. J Cell Biol. 143:195-205. Eyer, J., D.W. Cleveland, P.C. Wong, and A.C. Peterson. 1998. Pathogenesis of two axonopathies does not require axonal neurofilaments. Nature. 391:584-587. Eyer, J., and A. Peterson. 1994. Neurofilament-deficient axons and perikaryal aggregates in viable transgenic mice expressing a neurofilament-beta-galactosidase fusion protein. Neuron. 12:389-405. Fabrizi, G.M., T. Cavallaro, C. Angiari, L. Bertolasi, I. Cabrini, M. Ferrarini, and N. Rizzuto. 2004. Giant axon and neurofilament accumulation in Charcot-Marie- Tooth disease type 2E. Neurology. 62:1429-1431. Farrer, M.J., M.M. Hulihan, J.M. Kachergus, J.C. Dachsel, A.J. Stoessl, L.L. Grantier, S. Calne, D.B. Calne, B. Lechevalier, F. Chapon, Y. Tsuboi, T. Yamada, L. Gutmann, B. Elibol, K.P. Bhatia, C. Wider, C. Vilarino-Guell, O.A. Ross, L.A. Brown, M. Castanedes-Casey, D.W. Dickson, and Z.K. Wszolek. 2009. DCTN1 mutations in Perry syndrome. Nat Genet. 41:163-165. Fernyhough, P., A. Gallagher, S.A. Averill, J.V. Priestley, L. Hounsom, J. Patel, and D.R. Tomlinson. 1999. Aberrant neurofilament phosphorylation in sensory neurons of rats with diabetic neuropathy. Diabetes. 48:881-889. Fichera, M., M. Lo Giudice, M. Falco, M. Sturnio, S. Amata, O. Calabrese, S. Bigoni, E. Calzolari, and M. Neri. 2004. Evidence of kinesin heavy chain (KIF5A) involvement in pure hereditary spastic paraplegia. Neurology. 63:1108-1110. Francis, F., S. Roy, S.T. Brady, and M.M. Black. 2005. Transport of neurofilaments in growing axons requires microtubules but not actin filaments. J Neurosci Res. 79:442-450.

160

Galbraith, J.A., T.S. Reese, M.L. Schlief, and P.E. Gallant. 1999. Slow transport of unpolymerized tubulin and polymerized neurofilament in the squid giant axon. Proc Natl Acad Sci U S A. 96:11589-11594. Garcia, M.L., M.V. Rao, J. Fujimoto, V.B. Garcia, S.B. Shah, J. Crum, T. Gotow, Y. Uchiyama, M. Ellisman, N.A. Calcutt, and D.W. Cleveland. 2009. Phosphorylation of highly conserved neurofilament medium KSP repeats is not required for myelin-dependent radial axonal growth. J Neurosci. 29:1277-1284. Gee, M.A., J.E. Heuser, and R.B. Vallee. 1997. An extended microtubule-binding structure within the dynein motor domain. Nature. 390:636-639. Geisler, N., E. Kaufmann, S. Fischer, U. Plessmann, and K. Weber. 1983. Neurofilament architecture combines structural principles of intermediate filaments with carboxy-terminal extensions increasing in size between triplet proteins. EMBO J. 2:1295-1302. Georgiou, D.M., J. Zidar, M. Korosec, L.T. Middleton, T. Kyriakides, and K. Christodoulou. 2002. A novel NF-L mutation Pro22Ser is associated with CMT2 in a large Slovenian family. Neurogenetics. 4:93-96. Gill, S.R., P.C. Wong, M.J. Monteiro, and D.W. Cleveland. 1990. Assembly properties of dominant and recessive mutations in the small mouse neurofilament (NF-L) subunit. J Cell Biol. 111:2005-2019. Goldstein, L.S., and Z. Yang. 2000. Microtubule-based transport systems in neurons: the roles of kinesins and dyneins. Annu Rev Neurosci. 23:39-71. Grafstein, B., and D.S. Forman. 1980. Intracellular transport in neurons. Physiol Rev. 60:1167-1283. Gross, S.P. 2004. Hither and yon: a review of bi-directional microtubule-based transport. Phys Biol. 1:R1-11. Gu, C., W. Zhou, M.A. Puthenveedu, M. Xu, Y.N. Jan, and L.Y. Jan. 2006. The microtubule plus-end tracking protein EB1 is required for Kv1 voltage-gated K+ channel axonal targeting. Neuron. 52:803-816. Guidato, S., L.H. Tsai, J. Woodgett, and C.C. Miller. 1996. Differential cellular phosphorylation of neurofilament heavy side-arms by glycogen synthase kinase-3 and cyclin-dependent kinase-5. J Neurochem. 66:1698-1706. Hackney, D.D. 1995. Highly processive microtubule-stimulated ATP hydrolysis by dimeric kinesin head domains. Nature. 377:448-450.

161

Hafezparast, M., R. Klocke, C. Ruhrberg, A. Marquardt, A. Ahmad-Annuar, S. Bowen, G. Lalli, A.S. Witherden, H. Hummerich, S. Nicholson, P.J. Morgan, R. Oozageer, J.V. Priestley, S. Averill, V.R. King, S. Ball, J. Peters, T. Toda, A. Yamamoto, Y. Hiraoka, M. Augustin, D. Korthaus, S. Wattler, P. Wabnitz, C. Dickneite, S. Lampel, F. Boehme, G. Peraus, A. Popp, M. Rudelius, J. Schlegel, H. Fuchs, M. Hrabe de Angelis, G. Schiavo, D.T. Shima, A.P. Russ, G. Stumm, J.E. Martin, and E.M. Fisher. 2003. Mutations in dynein link motor neuron degeneration to defects in retrograde transport. Science. 300:808-812. Hancock, W.O., and J. Howard. 1998. Processivity of the motor protein kinesin requires two heads. J Cell Biol. 140:1395-1405. Harvey Lodish, A.B., S Lawrence Zipursky, Paul Matsudaira, David Baltimore, and James Darnell. 2000. Molecular Cell Biology. He, Y., F. Francis, K.A. Myers, W. Yu, M.M. Black, and P.W. Baas. 2005. Role of cytoplasmic dynein in the axonal transport of microtubules and neurofilaments. J Cell Biol. 168:697-703. Heidemann, S.R., J.M. Landers, and M.A. Hamborg. 1981. Polarity orientation of axonal microtubules. J Cell Biol. 91:661-665. Heins, S., P.C. Wong, S. Muller, K. Goldie, D.W. Cleveland, and U. Aebi. 1993. The rod domain of NF-L determines neurofilament architecture, whereas the end domains specify filament assembly and network formation. J Cell Biol. 123:1517-1533. Helfand, B.T., P. Loomis, M. Yoon, and R.D. Goldman. 2003. Rapid transport of neural intermediate filament protein. J Cell Sci. 116:2345-2359. Herrmann, H., H. Bar, L. Kreplak, S.V. Strelkov, and U. Aebi. 2007. Intermediate filaments: from cell architecture to nanomechanics. Nat Rev Mol Cell Biol. 8:562- 573. Hirai, T., M. Mizutani, K. Ochiai, T. Umemura, and C. Itakura. 1999. Distal axonopathy does not occur without neurofilament accumulation in gamma-diketone neuropathy: comparative studies of normal and neurofilament-deficient quail. Acta Neuropathol. 97:552-556. Hirano, A., H. Donnenfeld, S. Sasaki, and I. Nakano. 1984. Fine structural observations of neurofilamentous changes in amyotrophic lateral sclerosis. J Neuropathol Exp Neurol. 43:461-470. Hirokawa, N. 1998. Kinesin and dynein superfamily proteins and the mechanism of organelle transport. Science. 279:519-526. 162

Hirokawa, N., M.A. Glicksman, and M.B. Willard. 1984. Organization of mammalian neurofilament polypeptides within the neuronal cytoskeleton. J Cell Biol. 98:1523-1536. Hirokawa, N., S. Niwa, and Y. Tanaka. 2010. Molecular motors in neurons: transport mechanisms and roles in brain function, development, and disease. Neuron. 68:610-638. Hirokawa, N., Y. Noda, Y. Tanaka, and S. Niwa. 2009. Kinesin superfamily motor proteins and intracellular transport. Nat Rev Mol Cell Biol. 10:682-696. Hirokawa, N., and R. Takemura. 2005. Molecular motors and mechanisms of directional transport in neurons. Nat Rev Neurosci. 6:201-214. Hisanaga, S., Y. Gonda, M. Inagaki, A. Ikai, and N. Hirokawa. 1990. Effects of phosphorylation of the neurofilament L protein on filamentous structures. Cell Regul. 1:237-248. Hisanaga, S., and N. Hirokawa. 1988. Structure of the peripheral domains of neurofilaments revealed by low angle rotary shadowing. J Mol Biol. 202:297-305. Hisanaga, S., S. Yasugawa, T. Yamakawa, E. Miyamoto, M. Ikebe, M. Uchiyama, and T. Kishimoto. 1993. Dephosphorylation of microtubule-binding sites at the neurofilament-H tail domain by alkaline, acid, and protein phosphatases. J Biochem. 113:705-709. Hollenbeck, P.J. 1996. The pattern and mechanism of mitochondrial transport in axons. Front Biosci. 1:d91-102. Hook, P., and R.B. Vallee. 2006. The dynein family at a glance. J Cell Sci. 119:4369- 4371. Howard, J., A.J. Hudspeth, and R.D. Vale. 1989. Movement of microtubules by single kinesin molecules. Nature. 342:154-158. Hoyt, M.A., L. He, L. Totis, and W.S. Saunders. 1993. Loss of function of Saccharomyces cerevisiae kinesin-related CIN8 and KIP1 is suppressed by KAR3 motor domain mutations. Genetics. 135:35-44. Huang, J.D., S.T. Brady, B.W. Richards, D. Stenolen, J.H. Resau, N.G. Copeland, and N.A. Jenkins. 1999. Direct interaction of microtubule- and actin-based transport motors. Nature. 397:267-270. Ince, P.G., J.R. Highley, J. Kirby, S.B. Wharton, H. Takahashi, M.J. Strong, and P.J. Shaw. 2011. Molecular pathology and genetic advances in amyotrophic lateral

163

sclerosis: an emerging molecular pathway and the significance of glial pathology. Acta Neuropathol. Ishikawa, H., R. Bischoff, and H. Holtzer. 1968. Mitosis and intermediate-sized filaments in developing skeletal muscle. J Cell Biol. 38:538-555. Julien, J.P., and W.E. Mushynski. 1982. Multiple phosphorylation sites in mammalian neurofilament polypeptides. J Biol Chem. 257:10467-10470. Julien, J.P., and W.E. Mushynski. 1983. The distribution of phosphorylation sites among identified proteolytic fragments of mammalian neurofilaments. J Biol Chem. 258:4019-4025. Jung, C., T.M. Chylinski, A. Pimenta, D. Ortiz, and T.B. Shea. 2004. Neurofilament transport is dependent on actin and myosin. J Neurosci. 24:9486-9496. Jung, C., S. Lee, D. Ortiz, Q. Zhu, J.P. Julien, and T.B. Shea. 2005. The high and middle molecular weight neurofilament subunits regulate the association of neurofilaments with kinesin: inhibition by phosphorylation of the high molecular weight subunit. Brain Res Mol Brain Res. 141:151-155. Kaech, S., and G. Banker. 2006. Culturing hippocampal neurons. Nat Protoc. 1:2406- 2415. Karcher, R.L., S.W. Deacon, and V.I. Gelfand. 2002. Motor-cargo interactions: the key to transport specificity. Trends Cell Biol. 12:21-27. Karki, S., and E.L. Holzbaur. 1999. Cytoplasmic dynein and dynactin in and intracellular transport. Curr Opin Cell Biol. 11:45-53. Kasher, P.R., K.J. De Vos, S.B. Wharton, C. Manser, E.J. Bennett, M. Bingley, J.D. Wood, R. Milner, C.J. McDermott, C.C. Miller, P.J. Shaw, and A.J. Grierson. 2009. Direct evidence for axonal transport defects in a novel mouse model of mutant spastin-induced hereditary spastic paraplegia (HSP) and human HSP patients. J Neurochem. 110:34-44. King, S.J., and T.A. Schroer. 2000. Dynactin increases the processivity of the cytoplasmic dynein motor. Nat Cell Biol. 2:20-24. Knull, H.R., and J.L. Walsh. 1992. Association of glycolytic enzymes with the cytoskeleton. Curr Top Cell Regul. 33:15-30. Kong, J., V.W. Tung, J. Aghajanian, and Z. Xu. 1998. Antagonistic roles of neurofilament subunits NF-H and NF-M against NF-L in shaping dendritic arborization in spinal motor neurons. J Cell Biol. 140:1167-1176.

164

Kong, J., and Z. Xu. 2000. Overexpression of neurofilament subunit NF-L and NF-H extends survival of a mouse model for amyotrophic lateral sclerosis. Neurosci Lett. 281:72-74. Kreplak, L., H. Bar, J.F. Leterrier, H. Herrmann, and U. Aebi. 2005. Exploring the mechanical behavior of single intermediate filaments. J Mol Biol. 354:569-577. LaMonte, B.H., K.E. Wallace, B.A. Holloway, S.S. Shelly, J. Ascano, M. Tokito, T. Van Winkle, D.S. Howland, and E.L. Holzbaur. 2002. Disruption of dynein/dynactin inhibits axonal transport in motor neurons causing late-onset progressive degeneration. Neuron. 34:715-727. Lasek, R.J. 1986. Polymer sliding in axons. J Cell Sci Suppl. 5:161-179. Lasek, R.J., J.A. Garner, and S.T. Brady. 1984. Axonal transport of the cytoplasmic matrix. J Cell Biol. 99:212s-221s. Lawrence, C.J., R.K. Dawe, K.R. Christie, D.W. Cleveland, S.C. Dawson, S.A. Endow, L.S. Goldstein, H.V. Goodson, N. Hirokawa, J. Howard, R.L. Malmberg, J.R. McIntosh, H. Miki, T.J. Mitchison, Y. Okada, A.S. Reddy, W.M. Saxton, M. Schliwa, J.M. Scholey, R.D. Vale, C.E. Walczak, and L. Wordeman. 2004. A standardized kinesin nomenclature. J Cell Biol. 167:19-22. Lee, M.K., J.R. Marszalek, and D.W. Cleveland. 1994. A mutant neurofilament subunit causes massive, selective motor neuron death: implications for the pathogenesis of human motor neuron disease. Neuron. 13:975-988. Lee, M.K., Z. Xu, P.C. Wong, and D.W. Cleveland. 1993. Neurofilaments are obligate heteropolymers in vivo. J Cell Biol. 122:1337-1350. Leterrier, J.F., J. Kas, J. Hartwig, R. Vegners, and P.A. Janmey. 1996. Mechanical effects of neurofilament cross-bridges. Modulation by phosphorylation, lipids, and interactions with F-actin. J Biol Chem. 271:15687-15694. Levy, J.R., C.J. Sumner, J.P. Caviston, M.K. Tokito, S. Ranganathan, L.A. Ligon, K.E. Wallace, B.H. LaMonte, G.G. Harmison, I. Puls, K.H. Fischbeck, and E.L. Holzbaur. 2006. A motor neuron disease-associated mutation in p150Glued perturbs dynactin function and induces protein aggregation. J Cell Biol. 172:733- 745. Ligon, L.A., and O. Steward. 2000. Movement of mitochondria in the axons and dendrites of cultured hippocampal neurons. J Comp Neurol. 427:340-350. Lim, S.S., K.J. Edson, P.C. Letourneau, and G.G. Borisy. 1990. A test of microtubule translocation during neurite elongation. J Cell Biol. 111:123-130. 165

Lim, S.S., P.J. Sammak, and G.G. Borisy. 1989. Progressive and spatially differentiated stability of microtubules in developing neuronal cells. J Cell Biol. 109:253-263. Lobsiger, C.S., M.L. Garcia, C.M. Ward, and D.W. Cleveland. 2005. Altered axonal architecture by removal of the heavily phosphorylated neurofilament tail domains strongly slows superoxide dismutase 1 mutant-mediated ALS. Proc Natl Acad Sci U S A. 102:10351-10356. Lorenz, T., and M. Willard. 1978. Subcellular fractionation of intra-axonally transport polypeptides in the rabbit visual system. Proc Natl Acad Sci U S A. 75:505-509. Macioce, P., G. Gambara, M. Bernassola, L. Gaddini, P. Torreri, G. Macchia, C. Ramoni, M. Ceccarini, and T.C. Petrucci. 2003. Beta-dystrobrevin interacts directly with kinesin heavy chain in brain. J Cell Sci. 116:4847-4856. Macioce, P., N. Gandolfi, C.L. Leung, S.S. Chin, F. Malchiodi-Albedi, M. Ceccarini, T.C. Petrucci, and R.K. Liem. 1999. Characterization of NF-L and betaIISigma1- interaction in live cells. Exp Cell Res. 250:142-154. Marszalek, J.R., T.L. Williamson, M.K. Lee, Z. Xu, P.N. Hoffman, M.W. Becher, T.O. Crawford, and D.W. Cleveland. 1996. Neurofilament subunit NF-H modulates axonal diameter by selectively slowing neurofilament transport. J Cell Biol. 135:711-724. Medori, R., L. Autilio-Gambetti, S. Monaco, and P. Gambetti. 1985. Experimental diabetic neuropathy: impairment of slow transport with changes in axon cross- sectional area. Proc Natl Acad Sci U S A. 82:7716-7720. Medori, R., H. Jenich, L. Autilio-Gambetti, and P. Gambetti. 1988. Experimental diabetic neuropathy: similar changes of slow axonal transport and axonal size in different animal models. J Neurosci. 8:1814-1821. Meier, J., S. Couillard-Despres, H. Jacomy, C. Gravel, and J.P. Julien. 1999. Extra neurofilament NF-L subunits rescue motor neuron disease caused by overexpression of the human NF-H gene in mice. J Neuropathol Exp Neurol. 58:1099-1110. Mercer, J.A., P.K. Seperack, M.C. Strobel, N.G. Copeland, and N.A. Jenkins. 1991. Novel myosin heavy chain encoded by murine dilute coat colour locus. Nature. 349:709-713. Mersiyanova, I.V., A.V. Perepelov, A.V. Polyakov, V.F. Sitnikov, E.L. Dadali, R.B. Oparin, A.N. Petrin, and O.V. Evgrafov. 2000. A new variant of Charcot-Marie-

166

Tooth disease type 2 is probably the result of a mutation in the neurofilament- light gene. Am J Hum Genet. 67:37-46. Mikami, A., S.H. Tynan, T. Hama, K. Luby-Phelps, T. Saito, J.E. Crandall, J.C. Besharse, and R.B. Vallee. 2002. Molecular structure of cytoplasmic dynein 2 and its distribution in neuronal and ciliated cells. J Cell Sci. 115:4801-4808. Miki, H., M. Setou, K. Kaneshiro, and N. Hirokawa. 2001. All kinesin superfamily protein, KIF, genes in mouse and human. Proc Natl Acad Sci U S A. 98:7004- 7011. Miller, K.E., and M.P. Sheetz. 2004. Axonal mitochondrial transport and potential are correlated. J Cell Sci. 117:2791-2804. Miller, K.E., and M.P. Sheetz. 2006. Direct evidence for coherent low velocity axonal transport of mitochondria. J Cell Biol. 173:373-381. Mochida, S., H. Kobayashi, Y. Matsuda, Y. Yuda, K. Muramoto, and Y. Nonomura. 1994. Myosin II is involved in transmitter release at synapses formed between rat sympathetic neurons in culture. Neuron. 13:1131-1142. Monaco, S., L. Autilio-Gambetti, D. Zabel, and P. Gambetti. 1985. Giant axonal neuropathy: acceleration of neurofilament transport in optic axons. Proc Natl Acad Sci U S A. 82:920-924. Morris, R.L., and P.J. Hollenbeck. 1993. The regulation of bidirectional mitochondrial transport is coordinated with axonal outgrowth. J Cell Sci. 104 ( Pt 3):917-927. Muller, M.J., S. Klumpp, and R. Lipowsky. 2008. Tug-of-war as a cooperative mechanism for bidirectional cargo transport by molecular motors. Proc Natl Acad Sci U S A. 105:4609-4614. Muller, M.J., S. Klumpp, and R. Lipowsky. 2010. Bidirectional transport by molecular motors: enhanced processivity and response to external forces. Biophys J. 98:2610-2618. Munoz, D.G., C. Greene, D.P. Perl, and D.J. Selkoe. 1988. Accumulation of phosphorylated neurofilaments in anterior horn motoneurons of amyotrophic lateral sclerosis patients. J Neuropathol Exp Neurol. 47:9-18. Murayama, S., T.W. Bouldin, and K. Suzuki. 1992. Immunocytochemical and ultrastructural studies of upper motor neurons in amyotrophic lateral sclerosis. Acta Neuropathol. 83:518-524. Myers, K.A., and P.W. Baas. 2007. Kinesin-5 regulates the growth of the axon by acting as a brake on its microtubule array. J Cell Biol. 178:1081-1091. 167

Neuwald, A.F., L. Aravind, J.L. Spouge, and E.V. Koonin. 1999. AAA+: A class of chaperone-like ATPases associated with the assembly, operation, and disassembly of protein complexes. Genome Res. 9:27-43. Nguyen, M.D., R.C. Lariviere, and J.P. Julien. 2000. Reduction of axonal caliber does not alleviate motor neuron disease caused by mutant superoxide dismutase 1. Proc Natl Acad Sci U S A. 97:12306-12311. Nguyen, M.D., R.C. Lariviere, and J.P. Julien. 2001. Deregulation of Cdk5 in a mouse model of ALS: toxicity alleviated by perikaryal neurofilament inclusions. Neuron. 30:135-147. Nguyen, M.D., T. Shu, K. Sanada, R.C. Lariviere, H.C. Tseng, S.K. Park, J.P. Julien, and L.H. Tsai. 2004. A NUDEL-dependent mechanism of neurofilament assembly regulates the integrity of CNS neurons. Nat Cell Biol. 6:595-608. Nixon, R.A., and T.B. Shea. 1992. Dynamics of neuronal intermediate filaments: a developmental perspective. Cell Motil Cytoskeleton. 22:81-91. Ohara, O., Y. Gahara, T. Miyake, H. Teraoka, and T. Kitamura. 1993. Neurofilament deficiency in quail caused by nonsense mutation in neurofilament-L gene. J Cell Biol. 121:387-395. Okabe, S., and N. Hirokawa. 1990. Turnover of fluorescently labelled tubulin and actin in the axon. Nature. 343:479-482. Okabe, S., H. Miyasaka, and N. Hirokawa. 1993. Dynamics of the neuronal intermediate filaments. J Cell Biol. 121:375-386. Pachter, J.S., and R.K. Liem. 1984. The differential appearance of neurofilament triplet polypeptides in the developing rat optic nerve. Dev Biol. 103:200-210. Pant, H.C., and Veeranna. 1995. Neurofilament phosphorylation. Biochem Cell Biol. 73:575-592. Pastural, E., F.J. Barrat, R. Dufourcq-Lagelouse, S. Certain, O. Sanal, N. Jabado, R. Seger, C. Griscelli, A. Fischer, and G. de Saint Basile. 1997. Griscelli disease maps to 15q21 and is associated with mutations in the myosin-Va gene. Nat Genet. 16:289-292. Pazour, G.J., B.L. Dickert, and G.B. Witman. 1999. The DHC1b (DHC2) isoform of cytoplasmic dynein is required for flagellar assembly. J Cell Biol. 144:473-481. Peiffer, J., W. Schlote, A. Bischoff, E. Boltshauser, and G. Muller. 1977. Generalized giant axonal neuropathy: a filament-forming disease of neuronal, endothelial,

168

glial, and schwann cells in a patient without kinky hair. Acta Neuropathol. 40:213-218. Perez-Olle, R., M.A. Lopez-Toledano, D. Goryunov, N. Cabrera-Poch, L. Stefanis, K. Brown, and R.K. Liem. 2005. Mutations in the neurofilament light gene linked to Charcot-Marie-Tooth disease cause defects in transport. J Neurochem. 93:861- 874. Perrot, R., and J. Eyer. 2009. Neuronal intermediate filaments and neurodegenerative disorders. Brain Res Bull. 80:282-295. Pfister, K.K., P.R. Shah, H. Hummerich, A. Russ, J. Cotton, A.A. Annuar, S.M. King, and E.M. Fisher. 2006. Genetic analysis of the cytoplasmic dynein subunit families. PLoS Genet. 2:e1. Pfister, K.K., M.C. Wagner, D.L. Stenoien, S.T. Brady, and G.S. Bloom. 1989. Monoclonal antibodies to kinesin heavy and light chains stain vesicle-like structures, but not microtubules, in cultured cells. J Cell Biol. 108:1453-1463. Porter, M.E., R. Bower, J.A. Knott, P. Byrd, and W. Dentler. 1999. Cytoplasmic dynein heavy chain 1b is required for flagellar assembly in Chlamydomonas. Mol Biol Cell. 10:693-712. Prahlad, V., B.T. Helfand, G.M. Langford, R.D. Vale, and R.D. Goldman. 2000. Fast transport of neurofilament protein along microtubules in squid axoplasm. J Cell Sci. 113 ( Pt 22):3939-3946. Prahlad, V., M. Yoon, R.D. Moir, R.D. Vale, and R.D. Goldman. 1998. Rapid movements of vimentin on microtubule tracks: kinesin-dependent assembly of intermediate filament networks. J Cell Biol. 143:159-170. Puls, I., C. Jonnakuty, B.H. LaMonte, E.L. Holzbaur, M. Tokito, E. Mann, M.K. Floeter, K. Bidus, D. Drayna, S.J. Oh, R.H. Brown, Jr., C.L. Ludlow, and K.H. Fischbeck. 2003. Mutant dynactin in motor neuron disease. Nat Genet. 33:455-456. Puls, I., S.J. Oh, C.J. Sumner, K.E. Wallace, M.K. Floeter, E.A. Mann, W.R. Kennedy, G. Wendelschafer-Crabb, A. Vortmeyer, R. Powers, K. Finnegan, E.L. Holzbaur, K.H. Fischbeck, and C.L. Ludlow. 2005. Distal spinal and bulbar muscular atrophy caused by dynactin mutation. Ann Neurol. 57:687-694. Rahman, A., D.S. Friedman, and L.S. Goldstein. 1998. Two kinesin light chain genes in mice. Identification and characterization of the encoded proteins. J Biol Chem. 273:15395-15403.

169

Rao, M.V., J. Campbell, A. Yuan, A. Kumar, T. Gotow, Y. Uchiyama, and R.A. Nixon. 2003. The neurofilament middle molecular mass subunit carboxyl-terminal tail domains is essential for the radial growth and cytoskeletal architecture of axons but not for regulating neurofilament transport rate. J Cell Biol. 163:1021-1031. Rao, M.V., L.J. Engle, P.S. Mohan, A. Yuan, D. Qiu, A. Cataldo, L. Hassinger, S. Jacobsen, V.M. Lee, A. Andreadis, J.P. Julien, P.C. Bridgman, and R.A. Nixon. 2002a. Myosin Va binding to neurofilaments is essential for correct myosin Va distribution and transport and neurofilament density. J Cell Biol. 159:279-290. Rao, M.V., M.L. Garcia, Y. Miyazaki, T. Gotow, A. Yuan, S. Mattina, C.M. Ward, N.A. Calcutt, Y. Uchiyama, R.A. Nixon, and D.W. Cleveland. 2002b. Gene replacement in mice reveals that the heavily phosphorylated tail of neurofilament heavy subunit does not affect axonal caliber or the transit of cargoes in slow axonal transport. J Cell Biol. 158:681-693. Rao, M.V., M.K. Houseweart, T.L. Williamson, T.O. Crawford, J. Folmer, and D.W. Cleveland. 1998. Neurofilament-dependent radial growth of motor axons and axonal organization of neurofilaments does not require the neurofilament heavy subunit (NF-H) or its phosphorylation. J Cell Biol. 143:171-181. Rao, M.V., P.S. Mohan, A. Kumar, A. Yuan, L. Montagna, J. Campbell, Veeranna, E.M. Espreafico, J.P. Julien, and R.A. Nixon. 2011. The myosin Va head domain binds to the neurofilament-L rod and modulates endoplasmic reticulum (ER) content and distribution within axons. PLoS One. 6:e17087. Reid, E., M. Kloos, A. Ashley-Koch, L. Hughes, S. Bevan, I.K. Svenson, F.L. Graham, P.C. Gaskell, A. Dearlove, M.A. Pericak-Vance, D.C. Rubinsztein, and D.A. Marchuk. 2002. A kinesin heavy chain (KIF5A) mutation in hereditary spastic paraplegia (SPG10). Am J Hum Genet. 71:1189-1194. Roder, H.M., and V.M. Ingram. 1991. Two novel kinases phosphorylate tau and the KSP site of heavy neurofilament subunits in high stoichiometric ratios. J Neurosci. 11:3325-3343. Roy, J., S. Minotti, L. Dong, D.A. Figlewicz, and H.D. Durham. 1998. Glutamate potentiates the toxicity of mutant Cu/Zn-superoxide dismutase in motor neurons by postsynaptic calcium-dependent mechanisms. J Neurosci. 18:9673-9684. Roy, S., P. Coffee, G. Smith, R.K. Liem, S.T. Brady, and M.M. Black. 2000. Neurofilaments are transported rapidly but intermittently in axons: implications for slow axonal transport. J Neurosci. 20:6849-6861. 170

Ryu, J., L. Liu, T.P. Wong, D.C. Wu, A. Burette, R. Weinberg, Y.T. Wang, and M. Sheng. 2006. A critical role for myosin IIb in dendritic spine morphology and synaptic function. Neuron. 49:175-182. Sabry, J., T.P. O'Connor, and M.W. Kirschner. 1995. Axonal transport of tubulin in Ti1 pioneer neurons in situ. Neuron. 14:1247-1256. Salinas, S., C. Proukakis, A. Crosby, and T.T. Warner. 2008. Hereditary spastic paraplegia: clinical features and pathogenetic mechanisms. Lancet Neurol. 7:1127-1138. Samso, M., M. Radermacher, J. Frank, and M.P. Koonce. 1998. Structural characterization of a dynein motor domain. J Mol Biol. 276:927-937. Sanchez, I., L. Hassinger, R.K. Sihag, D.W. Cleveland, P. Mohan, and R.A. Nixon. 2000. Local control of neurofilament accumulation during radial growth of myelinating axons in vivo. Selective role of site-specific phosphorylation. J Cell Biol. 151:1013-1024. Sasaki, T., T. Gotow, M. Shiozaki, F. Sakaue, T. Saito, J.P. Julien, Y. Uchiyama, and S. Hisanaga. 2006. Aggregate formation and phosphorylation of neurofilament-L Pro22 Charcot-Marie-Tooth disease mutants. Hum Mol Genet. 15:943-952. Schliwa, M., and G. Woehlke. 2003. Molecular motors. Nature. 422:759-765. Schmidt, R.E., L.N. Beaudet, S.B. Plurad, and D.A. Dorsey. 1997. Axonal cytoskeletal pathology in aged and diabetic human sympathetic autonomic ganglia. Brain Res. 769:375-383. Scott, J.N., A.W. Clark, and D.W. Zochodne. 1999. Neurofilament and tubulin gene expression in progressive experimental diabetes: failure of synthesis and export by sensory neurons. Brain. 122 ( Pt 11):2109-2118. Sellers, J.R. 2000. Myosins: a diverse superfamily. Biochim Biophys Acta. 1496:3-22. Shah, J.V., L.A. Flanagan, P.A. Janmey, and J.F. Leterrier. 2000. Bidirectional translocation of neurofilaments along microtubules mediated in part by dynein/dynactin. Mol Biol Cell. 11:3495-3508. Shaw, G., and K. Weber. 1982. Differential expression of neurofilament triplet proteins in brain development. Nature. 298:277-279. Shetty, K.T., W.T. Link, and H.C. Pant. 1993. cdc2-like kinase from rat spinal cord specifically phosphorylates KSPXK motifs in neurofilament proteins: isolation and characterization. Proc Natl Acad Sci U S A. 90:6844-6848.

171

Shinoda, T., S. Taya, D. Tsuboi, T. Hikita, R. Matsuzawa, S. Kuroda, A. Iwamatsu, and K. Kaibuchi. 2007. DISC1 regulates neurotrophin-induced axon elongation via interaction with Grb2. J Neurosci. 27:4-14. Shubeita, G.T., S.L. Tran, J. Xu, M. Vershinin, S. Cermelli, S.L. Cotton, M.A. Welte, and S.P. Gross. 2008. Consequences of motor copy number on the intracellular transport of kinesin-1-driven lipid droplets. Cell. 135:1098-1107. Sihag, R.K., H. Jaffe, R.A. Nixon, and X. Rong. 1999. Serine-23 is a major protein kinase A phosphorylation site on the amino-terminal head domain of the middle molecular mass subunit of neurofilament proteins. J Neurochem. 72:491-499. Sihag, R.K., and R.A. Nixon. 1991. Identification of Ser-55 as a major protein kinase A phosphorylation site on the 70-kDa subunit of neurofilaments. Early turnover during axonal transport. J Biol Chem. 266:18861-18867. Skre, H. 1974. Genetic and clinical aspects of Charcot-Marie-Tooth's disease. Clin Genet. 6:98-118. Solecki, D.J., N. Trivedi, E.E. Govek, R.A. Kerekes, S.S. Gleason, and M.E. Hatten. 2009. Myosin II motors and F-actin dynamics drive the coordinated movement of the centrosome and soma during CNS glial-guided neuronal migration. Neuron. 63:63-80. Song, H., and S.A. Endow. 1998. Decoupling of nucleotide- and microtubule-binding sites in a kinesin mutant. Nature. 396:587-590. Stone, J.D., A.P. Peterson, J. Eyer, T.G. Oblak, and D.W. Sickles. 2001. Neurofilaments are nonessential to the pathogenesis of toxicant-induced axonal degeneration. J Neurosci. 21:2278-2287. Sun, D., C.L. Leung, and R.K. Liem. 1996. Phosphorylation of the high molecular weight neurofilament protein (NF-H) by Cdk5 and p35. J Biol Chem. 271:14245-14251. Szeverenyi, I., A.J. Cassidy, C.W. Chung, B.T. Lee, J.E. Common, S.C. Ogg, H. Chen, S.Y. Sim, W.L. Goh, K.W. Ng, J.A. Simpson, L.L. Chee, G.H. Eng, B. Li, D.P. Lunny, D. Chuon, A. Venkatesh, K.H. Khoo, W.H. McLean, Y.P. Lim, and E.B. Lane. 2008. The Human Intermediate Filament Database: comprehensive information on a gene family involved in many human diseases. Hum Mutat. 29:351-360. Tabb, J.S., B.J. Molyneaux, D.L. Cohen, S.A. Kuznetsov, and G.M. Langford. 1998. Transport of ER vesicles on actin filaments in neurons by myosin V. J Cell Sci. 111 ( Pt 21):3221-3234. 172

Takeda, S., T. Funakoshi, and N. Hirokawa. 1995. Tubulin dynamics in neuronal axons of living zebrafish embryos. Neuron. 14:1257-1264. Takeda, S., S. Okabe, T. Funakoshi, and N. Hirokawa. 1994. Differential dynamics of neurofilament-H protein and neurofilament-L protein in neurons. J Cell Biol. 127:173-185. Tanaka, M., and W. Herr. 1990. Differential transcriptional activation by Oct-1 and Oct- 2: interdependent activation domains induce Oct-2 phosphorylation. Cell. 60:375- 386. Tarrade, A., C. Fassier, S. Courageot, D. Charvin, J. Vitte, L. Peris, A. Thorel, E. Mouisel, N. Fonknechten, N. Roblot, D. Seilhean, A. Dierich, J.J. Hauw, and J. Melki. 2006. A mutation of spastin is responsible for swellings and impairment of transport in a region of axon characterized by changes in microtubule composition. Hum Mol Genet. 15:3544-3558. Taya, S., T. Shinoda, D. Tsuboi, J. Asaki, K. Nagai, T. Hikita, S. Kuroda, K. Kuroda, M. Shimizu, S. Hirotsune, A. Iwamatsu, and K. Kaibuchi. 2007. DISC1 regulates the transport of the NUDEL/LIS1/14-3-3epsilon complex through kinesin-1. J Neurosci. 27:15-26. Terada, S., T. Nakata, A.C. Peterson, and N. Hirokawa. 1996. Visualization of slow axonal transport in vivo. Science. 273:784-788. Theiss, C., M. Napirei, and K. Meller. 2005. Impairment of anterograde and retrograde neurofilament transport after anti-kinesin and anti-dynein antibody microinjection in chicken dorsal root ganglia. Eur J Cell Biol. 84:29-43. Thorn, K.S., J.A. Ubersax, and R.D. Vale. 2000. Engineering the processive run length of the kinesin motor. J Cell Biol. 151:1093-1100. Thyagarajan, A., M.J. Strong, and B.G. Szaro. 2007. Post-transcriptional control of neurofilaments in development and disease. Exp Cell Res. 313:2088-2097. Trivedi, N., P. Jung, and A. Brown. 2007. Neurofilaments switch between distinct mobile and stationary states during their transport along axons. J Neurosci. 27:507-516. Trojanowski, J.Q., M.L. Kelsten, and V.M. Lee. 1989. Phosphate-dependent and independent neurofilament protein epitopes are expressed throughout the in human medulloblastoma (D283 MED) cells. Am J Pathol. 135:747-758. Tu, P.H., P. Raju, K.A. Robinson, M.E. Gurney, J.Q. Trojanowski, and V.M. Lee. 1996. Transgenic mice carrying a human mutant superoxide dismutase transgene

173

develop neuronal cytoskeletal pathology resembling human amyotrophic lateral sclerosis lesions. Proc Natl Acad Sci U S A. 93:3155-3160. Twelvetrees, A.E., E.Y. Yuen, I.L. Arancibia-Carcamo, A.F. MacAskill, P. Rostaing, M.J. Lumb, S. Humbert, A. Triller, F. Saudou, Z. Yan, and J.T. Kittler. 2010. Delivery of GABAARs to synapses is mediated by HAP1-KIF5 and disrupted by mutant huntingtin. Neuron. 65:53-65. Uchida, A., N.H. Alami, and A. Brown. 2009. Tight functional coupling of kinesin-1A and dynein motors in the bidirectional transport of neurofilaments. Mol Biol Cell. 20:4997-5006. Uchida, A., and A. Brown. 2004. Arrival, reversal, and departure of neurofilaments at the tips of growing axons. Mol Biol Cell. 15:4215-4225. Vale, R.D. 2003. The toolbox for intracellular transport. Cell. 112:467- 480. Vale, R.D., T. Funatsu, D.W. Pierce, L. Romberg, Y. Harada, and T. Yanagida. 1996. Direct observation of single kinesin molecules moving along microtubules. Nature. 380:451-453. Vale, R.D., and R.A. Milligan. 2000. The way things move: looking under the hood of molecular motor proteins. Science. 288:88-95. Verhey, K.J., N. Kaul, and V. Soppina. 2011. Kinesin assembly and movement in cells. Annu Rev Biophys. 40:267-288. Vogel, P., M. Gabriel, H.H. Goebel, and P.J. Dyck. 1985. Hereditary motor sensory neuropathy type II with neurofilament accumulation: new finding or new disorder? Ann Neurol. 17:455-461. Wagner, O.I., J. Ascano, M. Tokito, J.F. Leterrier, P.A. Janmey, and E.L. Holzbaur. 2004. The interaction of neurofilaments with the microtubule motor cytoplasmic dynein. Mol Biol Cell. 15:5092-5100. Wang, L., and A. Brown. 2001. Rapid intermittent movement of axonal neurofilaments observed by fluorescence photobleaching. Mol Biol Cell. 12:3257-3267. Wang, L., and A. Brown. 2010. A hereditary spastic paraplegia mutation in kinesin- 1A/KIF5A disrupts neurofilament transport. Mol Neurodegener. 5:52. Wang, L., C.L. Ho, D. Sun, R.K. Liem, and A. Brown. 2000. Rapid movement of axonal neurofilaments interrupted by prolonged pauses. Nat Cell Biol. 2:137-141.

174

Wang, W., J. Ding, E. Allen, P. Zhu, L. Zhang, H. Vogel, and Y. Yang. 2005. Gigaxonin interacts with tubulin folding cofactor B and controls its degradation through the ubiquitin-proteasome pathway. Curr Biol. 15:2050-2055. Weiss, P., and H.B. Hiscoe. 1948. Experiments on the mechanism of nerve growth. J Exp Zool. 107:315-395. Welte, M.A. 2004. Bidirectional transport along microtubules. Curr Biol. 14:R525-537. Welzel, O., J. Knorr, A.M. Stroebel, J. Kornhuber, and T.W. Groemer. 2011. A fast and robust method for automated analysis of axonal transport. Eur Biophys J. 40:1061-1069. Willard, M., and C. Simon. 1983. Modulations of neurofilament axonal transport during the development of rabbit retinal ganglion cells. Cell. 35:551-559. Williamson, T.L., L.I. Bruijn, Q. Zhu, K.L. Anderson, S.D. Anderson, J.P. Julien, and D.W. Cleveland. 1998. Absence of neurofilaments reduces the selective vulnerability of motor neurons and slows disease caused by a familial amyotrophic lateral sclerosis-linked superoxide dismutase 1 mutant. Proc Natl Acad Sci U S A. 95:9631-9636. Williamson, T.L., and D.W. Cleveland. 1999. Slowing of axonal transport is a very early event in the toxicity of ALS-linked SOD1 mutants to motor neurons. Nat Neurosci. 2:50-56. Wong, P.C., and D.W. Cleveland. 1990. Characterization of dominant and recessive assembly-defective mutations in mouse neurofilament NF-M. J Cell Biol. 111:1987-2003. Wong, P.C., J. Marszalek, T.O. Crawford, Z. Xu, S.T. Hsieh, J.W. Griffin, and D.W. Cleveland. 1995. Increasing neurofilament subunit NF-M expression reduces axonal NF-H, inhibits radial growth, and results in neurofilamentous accumulation in motor neurons. J Cell Biol. 130:1413-1422. Wong, Y.L., and S.E. Rice. 2010. Kinesin's light chains inhibit the head- and microtubule-binding activity of its tail. Proc Natl Acad Sci U S A. 107:11781- 11786. Xia, C., A. Rahman, Z. Yang, and L.S. Goldstein. 1998. Chromosomal localization reveals three kinesin heavy chain genes in mouse. Genomics. 52:209-213. Xia, C.H., E.A. Roberts, L.S. Her, X. Liu, D.S. Williams, D.W. Cleveland, and L.S. Goldstein. 2003. Abnormal neurofilament transport caused by targeted disruption of neuronal kinesin heavy chain KIF5A. J Cell Biol. 161:55-66. 175

Xu, Z., L.C. Cork, J.W. Griffin, and D.W. Cleveland. 1993. Increased expression of neurofilament subunit NF-L produces morphological alterations that resemble the pathology of human motor neuron disease. Cell. 73:23-33. Yabe, J.T., W.K. Chan, T.M. Chylinski, S. Lee, A.F. Pimenta, and T.B. Shea. 2001a. The predominant form in which neurofilament subunits undergo axonal transport varies during axonal initiation, elongation, and maturation. Cell Motil Cytoskeleton. 48:61-83. Yabe, J.T., T. Chylinski, F.S. Wang, A. Pimenta, S.D. Kattar, M.D. Linsley, W.K. Chan, and T.B. Shea. 2001b. Neurofilaments consist of distinct populations that can be distinguished by C-terminal phosphorylation, bundling, and axonal transport rate in growing axonal neurites. J Neurosci. 21:2195-2205. Yabe, J.T., C. Jung, W.K. Chan, and T.B. Shea. 2000. Phospho-dependent association of neurofilament proteins with kinesin in situ. Cell Motil Cytoskeleton. 45:249-262. Yabe, J.T., A. Pimenta, and T.B. Shea. 1999. Kinesin-mediated transport of neurofilament protein oligomers in growing axons. J Cell Sci. 112 ( Pt 21):3799- 3814. Yamasaki, H., G.S. Bennett, C. Itakura, and M. Mizutani. 1992. Defective expression of neurofilament protein subunits in hereditary hypotrophic axonopathy of quail. Lab Invest. 66:734-743. Yamasaki, H., C. Itakura, and M. Mizutani. 1991. Hereditary hypotrophic axonopathy with neurofilament deficiency in a mutant strain of the Japanese quail. Acta Neuropathol. 82:427-434. Yan, Y., and A. Brown. 2005. Neurofilament polymer transport in axons. J Neurosci. 25:7014-7021. Yan, Y., K. Jensen, and A. Brown. 2007. The polypeptide composition of moving and stationary neurofilaments in cultured sympathetic neurons. Cell Motil Cytoskeleton. 64:299-309. Young, E.C., H.K. Mahtani, and J. Gelles. 1998. One-headed kinesin derivatives move by a nonprocessive, low-duty ratio mechanism unlike that of two-headed kinesin. Biochemistry. 37:3467-3479. Yuan, A., M.V. Rao, T. Sasaki, Y. Chen, A. Kumar, Veeranna, R.K. Liem, J. Eyer, A.C. Peterson, J.P. Julien, and R.A. Nixon. 2006. Alpha-internexin is structurally and functionally associated with the neurofilament triplet proteins in the mature CNS. J Neurosci. 26:10006-10019. 176

Zhang, B., P. Tu, F. Abtahian, J.Q. Trojanowski, and V.M. Lee. 1997. Neurofilaments and orthograde transport are reduced in ventral root axons of transgenic mice that express human SOD1 with a G93A mutation. J Cell Biol. 139:1307-1315. Zhang, K., Y. Osakada, W. Xie, and B. Cui. 2011. Automated image analysis for tracking cargo transport in axons. Microsc Res Tech. 74:605-613. Zhang, Z., D.M. Casey, J.P. Julien, and Z. Xu. 2002. Normal dendritic arborization in spinal motoneurons requires neurofilament subunit L. J Comp Neurol. 450:144- 152. Zhao, C., J. Takita, Y. Tanaka, M. Setou, T. Nakagawa, S. Takeda, H.W. Yang, S. Terada, T. Nakata, Y. Takei, M. Saito, S. Tsuji, Y. Hayashi, and N. Hirokawa. 2001. Charcot-Marie-Tooth disease type 2A caused by mutation in a microtubule motor KIF1Bbeta. Cell. 105:587-597. Zhou, H.M., I. Brust-Mascher, and J.M. Scholey. 2001. Direct visualization of the movement of the monomeric axonal transport motor UNC-104 along neuronal processes in living Caenorhabditis elegans. J Neurosci. 21:3749-3755. Zhu, Q., S. Couillard-Despres, and J.P. Julien. 1997. Delayed maturation of regenerating myelinated axons in mice lacking neurofilaments. Exp Neurol. 148:299-316. Zhu, Q., M. Lindenbaum, F. Levavasseur, H. Jacomy, and J.P. Julien. 1998. Disruption of the NF-H gene increases axonal microtubule content and velocity of neurofilament transport: relief of axonopathy resulting from the toxin beta,beta'- iminodipropionitrile. J Cell Biol. 143:183-193. Zochodne, D.W., H.S. Sun, C. Cheng, and J. Eyer. 2004. Accelerated diabetic neuropathy in axons without neurofilaments. Brain. 127:2193-2200.

177