THE ROLE OF VA AND THE / COMPLEX IN AXONAL 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:

Nael H Alami

Graduate Program in

Molecular, Cellular, and Developmental Biology

The Ohio State University

2009

Dissertation Committee:

Anthony Brown, Advisor

Harold Fisk

James Jontes

Dale Vandre

Copyright by

Nael H Alami

ABSTRACT

Neurofilaments are the major cytoskeletal elements in mature neuronal cells.

They are known for their space-filling properties and for forming an elastic network along the axons that is responsible for radial growth and maintaining proper caliber. and other cytoskeletal polymers, membranous organelles, and macromolecular cargo are transported along the axon in one of two directions: away from the cell body, towards the axon tips, in an anterograde fashion, or back towards the cell body from the direction of the growth cones, in a retrograde fashion. The regulation of this transport is vital for the functional and structural well-being of the neuron and is mainly dependent on the - based motor and dynein/dynactin.

In 2002, a study by Rao et al. suggested a role for the -based motor myosin Va in neurofilament transport. They reported that myosin Va associates with neurofilaments in vivo and that neurofilaments accumulate in axons of neurons lacking myosin Va. Based on these observations, we hypothesize that myosin Va is involved in neurofilament transport and that in the absence of myosin Va, neurofilaments move less efficiently along the axons. To test this hypothesis, we used fluorescent live-cell imaging of neurofilament movement in

ii SCG neurons from wild type and dilute lethal mice. Our results indicate that the absence of myosin Va from SCG neurons does not significantly alter neurofilament velocity or frequency of movement. We also used a fluorescence photoactivation pulse-escape technique to measure the rate of departure of photoactivatable GFP-tagged neurofilaments from photoactivated axonal regions in cultured DRG neurons from two strains of dilute lethal mice. We observed a

48%-169% increase in the mean time for neurofilaments to depart the activated regions in neurons from dilute lethal as compared to wild type. We conclude that neurofilaments pause for more prolonged periods in the absence of myosin Va.

We propose that myosin Va is a short-range motor for neurofilaments and that it can function to enhance the efficiency of neurofilament transport in axons by delivering neurofilaments to their microtubule tracks.

We also studied the role of dynein/dynactin in neurofilament transport.

Dynein/dynactin is a retrograde motor complex that was found to associate with neurofilaments in vivo and in vitro. It has been previously proposed that it is responsible for retrograde neurofilament transport but without any direct evidence. We used SCG and cortical neuronal cultures to observe neurofilament transport in cells where dynein/dynactin activity has been disrupted using a number of different approaches that target different subunits of the complex.

Using dynein heavy chain knock-down, dynein intermediate chain functional blocking antibody, dynamitin/p50 overexpression in SCG neurons and p150- coiled-coil1 overexpression in cortical neurons, we report an inhibition of

iii retrograde transport. This clearly indicates that dynein/dynactin is indeed the retrograde neurofilament motor. We also observed a reciprocal inhibition of anterograde transport that mirrored the retrograde transport inhibition in every one of these manipulations. This suggests that a tight functional coupling exists between the retrograde and anterograde motors of neurofilaments, where the activity of one motor is needed for the activity of the other and vice versa. In one of our observations after disrupting dynein/dynactin activity using p150-coiled-coil overexpression in SCG neurons, we report an increase in anterograde transport.

Our attempts to reproduce this result in cortical neurons, at different times after transfection, or with different transfection concentrations failed, as we observe an inhibition in both directions of transport in all such cases. This unique result, therefore, remains to be explored.

In conclusion, we propose that the transport and organization of neurofilaments may be orchestrated by the coordinated activity of at least three different motor proteins, kinesins, dynein/dynactin, and myosin-Va, which act together to convey and distribute these polymers along neuronal axons, and the disruption of any of these motors could lead to neurofilament transport defects and accumulations that could ultimately result in neuronal degeneration.

iv

To my family

v

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor Dr. Anthony Brown for his invaluable advice and support. The fruits of my work are a result of his supervision and guidance, and for that I am grateful.

I would also like to thank my committee members, professors Harold Fisk, James

Jontes and Dale Vandre for their patience, insightful suggestions and critical review of my dissertation.

Past and present members of the Brown lab have helped make the work environment throughout those past years enjoyable and exciting. I would like to thank Dr. Niraj Trivedi for his friendship and continuous support, Dr. Atsuko

Uchida, Lina Wang, Paula Monsma and Gulsen Colakoglu for their insightful critiques and encouragement when needed.

Throughout my stay in Columbus I have been blessed by the presence of a group of very special and dear friends, whose support and love I will always cherish and treasure. You have made my journey profoundly enjoyable and memorable, sharing with me the good as well as the bad times. My thanks and gratitude go out to all of you, especially to Niraj, Alice, Erica, Nadine, Sleiman,

Rami, Ihab, the lovely Nohal, my wonderful and inspirational friend Nesrine, my eternal friend Sarine, the beautiful Noura, and my second sister Zeina.

vi Last but not least, this and everything I have and will accomplish, I owe to three people who have made me the person that I am: my father, whose every step has been a guiding light along the way; my mother, whose unconditional love inspires me to become a better person; and my talented, gentle and loving sister,

Nadine. Your trust, belief, and unwavering support instill in me the strength and determination to carry on through all obstacles and hardship. To you, I am eternally grateful and indebted.

vii VITA

April 24, 1981………………………...…….Born- Mimess, Lebanon

June 2001…………………………..………Bachelor of Science in Biology

The American University of Beirut

Beirut, Lebanon

June, 2003…………………………………Master of Science in Biology

The American University of Beirut

Beirut, Lebanon

September 2003-Present………………..PhD Candidate, Molecular, Cellular and

Developmental Biology Graduate

Program, The Ohio State University,

Ohio, USA

PUBLICATIONS

Uchida A, Alami NH, Brown A. Tight functional coupling of kinein-1A and dynein motors in the bidirectional transport of neurofilaments. Mol Biol Cell. In press.

Alami NH, Brown A. Myosin Va increases the efficiency of neurofilament transport by decreasing the duration of long-term pausing. J Neuroscience. 2009 May 20; 29(20):6625-34.

viii

FIELD OF STUDY

Major Field: Molecular, Cellular and Developmental Biology

ix

TABLE OF CONTENTS

Abstract ...... ii

Acknowledgements ...... vi

Vita ...... viii

List of Tables ...... xiii

List of Figures ...... xiv

Chapter 1: INTRODUCTION ...... 1

1.1. Neurofilaments...... 1 1.1.1. General introduction ...... 1 1.1.2. Neurofilament organization and assembly ...... 5 1.1.3. Neurofilament function ...... 8

1.2. Axonal transport...... 13 1.2.1. A historical perspective ...... 13 1.2.2. Fast axonal transport...... 15 1.2.3. Slow axonal transport...... 16 1.2.4. Polymers vs monomers...... 18 1.2.5. Neurofilament phosphorylation and transport ...... 19

1.3. Neurofilaments and neurodegenerative disease ...... 22 1.3.1. Amyotrophic lateral sclerosis...... 23 1.3.2. Alzheimer’s disease ...... 25 1.3.3. Parkinson’s disease ...... 26 1.3.4. Charcot-Marie-Tooth disease...... 26 1.3.5. Neuronal inclusion disease ...... 28 1.3.6. Diabetic neuropathy ...... 29 1.3.7. Giant axonal neuropathy ...... 30

1.4. Molecular motors ...... 30 1.4.1. superfamily ...... 32 1.4.2. The dynein/dynactin complex...... 35 1.4.3. The myosin superfamily ...... 40

Chapter 2: Materials and Methods ...... 46

2.1. Mice ...... 46

x 2.1.1. ICR mice...... 46 2.1.2. DLS/LeJ mice ...... 46 2.1.3. dl20J mice ...... 48

2.2. Cell culture ...... 50 2.2.1. SCG neuronal culture...... 50 2.2.2. DRG neuronal culture...... 50 2.2.3. Astroglial cell culture ...... 51 2.2.4. Cortical neuronal culture ...... 51

2.3. Cloning and transfection ...... 52 2.3.1. DNA constructs ...... 52 2.3.2. RNA interference...... 54 2.3.3. DNA and RNA injections ...... 55 2.3.4. Antibody injections ...... 56 2.3.5. Electroporation ...... 56

2.4. Live cell imaging ...... 57 2.4.1. Heated chamber system ...... 57 2.4.2. Neurofilament movement through gaps ...... 58 2.4.3. Fluorescence activation pulse-escape ...... 60 2.4.4. Calculation of mean time to depart...... 61 2.4.5. Statistical Analysis...... 63

2.5. Immunostaining...... 64

2.6. Western Blotting...... 65

Chapter 3: The Role of Myosin Va in Neurofilament Transport...... 67

3.1. Introduction ...... 67

3.2. Identification of dilute lethal mice from DLS/LeJ and dl20J strains...... 69

3.3. Neurofilament distribution along neuronal axons of dilute lethal mice ...... 71

3.4. Short-term mobile behavior of neurofilaments in SCG neurons ...... 72 3.4.1. Neurofilament length and frequency of movement in naturally occurring gaps...... 76 3.4.2. Kinetics of neurofilament transport along mouse SCG neurons ...... 77

3.5. Long-term mobile behavior of neurofilaments in DRG neurons...... 82 3.5.1. Long-term pausing in DRG neurons from DLS/LeJ mice ...... 83 3.5.2. Statistical analysis of pulse-escape data from DLS/LeJ mouse neurons ...... 86 3.5.3. Long-term pausing in DRG neurons from dl20J mice ...... 86

xi

3.6. Investigating neurofilament phosphorylation state in dilute lethal mice ....90

3.7. Summary...... 92

Chapter 4: The Role of the Dynein/Dynactin Complex in Neurofilament Transport ...... 94

4.1. Introduction ...... 94

4.2. Knock-down of dynein heavy chain in SCG neuronal cultures ...... 98

4.3. Functional inhibition of DIC in SCG neurons ...... 101

4.4. Dynamitin/p50 overexpression in SCG neurons ...... 103

4.5. Overexpression of coiled coil 1 domain of p150 in SCG neurons...... 107

4.6. Overexpression of coiled coil 1 domain of p150 in cortical neurons ...... 110

4.7. Summary...... 113

Chapter 5: Discussion ...... 116

5.1. Role of myosin Va in short-range transport ...... 116

5.2. The role of dynein/dynactin in neurofilament transport ...... 124 5.2.1. A simple tug-of-war model...... 127 5.2.2. The exclusionary presence model...... 128 5.2.3. The coordination model...... 129 5.2.4. The coordinated tug-of-war model ...... 130 5.2.5. Possible interpretations of CC1-p150 overexpression data ...... 131

5.3. Future directions ...... 132

5.4. A general model for neurofilament axonal transport ...... 135

xii

LIST OF TABLES

Table 1.1. Structural and mechanical characteristics of various myosin

classes…………………………………………………………………………….41

Table 2.1. Phenotypic identification of DLS/LeJ strain ...... 48

Table 3.1. Statistical comparison of neurofilament movement in wild type and dilute lethal neurons ...... 77

Table 4.1. Neurofilament transport along SCG neuronal axons in culture ...... 100

Table 4.2. Summary of statistical analysis for neurofilament movement in dynein inhibition experimets...... 101

Table 4.3. Neurofilament transport along SCG neuronal axons in culture: control vs p50...... 106

Table 4.4. Neurofilament transport along SCG neuronal axons in culture: control vs p150-CC1 ...... 109

Table 4.5. Neurofilament transport along cortical neurons in culture ...... 111

xiii

LIST OF FIGURES

Figure 1.1. Neurofilament protein organization and structure...... 4

Figure 1.2. A model for neurofilament assembly...... 8

Figure 1.3. Principal members of kinesin superfamily proteins...... 32

Figure 1.4. The dynein/dynactin complex...... 37

Figure 1.5. Schematic representation of three classes of myosin…………...... 39

Figure 1.6. Increased neurofilament content and density in dilute lethal axons.43

Figure 2.1. PCR genotyping of dl20J mice...... 48

Figure 3.1. Phenotyping and genotyping dilute lethal pups...... 73

Figure 3.2. Dilute lethal neurons extend axons that contain neurofilaments and gaps...... 73

Figure 3.3. Examples of moving neurofilaments in wild type and dilute lethal axons...... 74

Figure 3.4. Analysis of neurofilament movement in wild type and dilute lethal axons...... 79

Figure 3.5. A pulse-escape fluorescent activation experiment...... 83

Figure 3.6. Pulse-escape kinetics for wild type and DLS/LeJ mice...... 86

Figure 3.7. Pulse-escape kinetics for wild type and dl20J mice...... 87

Figure 3.8. Neurofilament phosphorylation state in wild type and dl20J mice...89

Figure 4.1. Efficiency of DHC siRNA knock down after 4 and 7 days in culture.97

xiv Figure 4.2. Effect of dynein heavy chain knock down on neurofilament movement...... 98

Figure 4.3. Effect of function-blocking dynein intermediate chain antibody (74.1) on neurofilament transport ...... 103

Figure 4.4. Summary of results after p50 overexpression in SCG neuronal cells ...... 107

Figure 4.5. Frequency of neurofilament transport after p150-CC1 overexpression in SCG neurons...... 109

Figure 4.6. Frequency of neurofilament transport after p150-CC1 overexpression in cortical neurons ...... 112

Figure 5.1. A schematic of the architecture of the subaxolemmal of a squid giant axon ...... 121

Figure 5.2. A schematic showing actin filaments as short-distance tracks for organelle transport...... 118

Figure 5.3. Myosin Va as a short-distance motor for neurofilament movement119

Figure 5.4. Three possible models for bidirectional transport...... 125

Figure 5.5. A unified model for neurofilament transport...... 135

xv

Chapter 1: INTRODUCTION

1.1. Neurofilaments

1.1.1. General introduction Neurofilaments, initially named neurofibrils, were first described in the nineteenth century by Valentin (1836), Purkinje (1838), Schulze (1871), and Cajal (1899), as fibrous networks within neurons. Neurofilaments are one of the most abundant cytoskeletal proteins in neurons and play an important role in maintaining the proper axonal caliber, which determines the rate of nerve impulse conduction.

They also convey mechanical strength on axons and dendrites, and function as spacers in these processes.

Neurofilaments are members of the intermediate filament family. The name

‘intermediate’ derives from their characteristic 8-10 nm-diameter, which is intermediate between the 6 nm-diameter of actin, and 24 nm-diameter of . Members of the intermediate filament superfamily are classified into six different classes (I-VI) based on the molecular structure homology of their

α-helical domain. These can be further divided into epithelial and non-epithelial groups based on their cellular localization. Types I and II filaments are found in

1 the epithelia, and include acidic (44-60 kDa; type I) and basic keratins

(50-70 kDa; type II). Type III intermediate filaments include , glial fibrillary acidic protein (GFAP), , and . Peripherin is a neuron specific intermediate filament protein (Portier et al., 1983; Leonard et al., 1988; Parysek et al., 1988; Greene, 1989). Vimentin is mainly found in non-neurons in adults, but it is also expressed in neuroblasts and certain unusual neurons (Drager,

1983; Shaw and Weber, 1983, 1984; Schwob et al., 1986).

Type IV intermediate filament proteins include the three major neurofilament subunits: NFL (low molecular weight, 66 kD), NFM (medium molecular weight,

95-100 kD) and NFH (high molecular weight, 110-115 kD) (Lee and Cleveland,

1996). These triplet proteins share similarity in protein and genomic sequence with α-, which is also neuron specific (Chiu et al., 1989; Kaplan et al.,

1990).

Nuclear , which are part of the nuclear lamina architecture (Aebi et al.,

1986; McKeon et al., 1986; Steinert and Roop, 1988), are type V intermediate filaments. is defined as a type VI intermediate filament (Dahlstrand et al.,

1992a; Dahlstrand et al., 1992b) and is found in neural stem cells and certain muscle tissue (Lendahl et al., 1990).

Like other intermediate filament proteins, neurofilament triplet proteins share an

N-terminal globular head domain, α-helical coiled coil rod domain and C-terminal

2 tail domain. The head domain of neurofilaments is rich in serines and threonines. Phosphorylation and O-glycosylation of these residues is believed to be important for regulation of neurofilament assembly. The rod domain in NFM and NFH is expected to form a single continuous coiled coil, which, in NFL, is interrupted by a heptad repeat. The diversity of neurofilament triplet proteins resides primarily in the carboxy-terminal tail domain. For NFL, this region contains many glutamate residues comprising a segment referred to as the “E” segment. Unlike NFL, NFM and NFH have much longer tail domains. In addition to E segments, they contain numerous lysine-serine-proline (KSP) repeats, with

42-51 in NFH, depending on the species, and much fewer in NFM (Napolitano et al., 1987; Shetty et al., 1993). The serines in these KSP domains are heavily phosphorylated in axons (Julien and Mushynski, 1982, 1983; Carden et al., 1985;

Lee et al., 1988b; Xu et al., 1992; Elhanany et al., 1994); Figure 1.1). It has been suggested that phosphorylation of the tail domains of NFM and NFH is involved in the elongation of neurofilaments and regulation of axonal calibers and neurofilament transport (Grant and Pant, 2000).

3

A.

B.

Figure 1.1. Neurofilament protein organization and structure

(A) an electron micrograph from frog axon showing a longitudinal view of a quick-freeze deep- etch preparation of neurofilament polymers running in parallel with each other at a distance of ~20-30 nm apart. Numerous cross-linkers (diameter 4-6 nm; length 20-30 nm) connect adjacent polymers like ladders. Scale bar: 0.1 µm x 203,000 (Hirokawa, 1982). (B) Schematic representation of mouse, rat, and human neurofilament subunits. The three subunits share a highly conserved α-helical domain of ~310 amino acids that is flanked by non-α-helical amino- (head) and carboxy-(tail) terminal end domains. The tail domains are variable in size. KSP repeats on NFM and NFH tails are heavily phosphorylated (Perrot et al., 2008).

4

1.1.2. Neurofilament organization and assembly

Neurofilaments are the main cytoskeletal elements in mature neurons making up

13% of total proteins and 50% of the Triton-soluble protein fraction (Morris and

Lasek, 1982). In the central nervous system, the neurofilament array is formed by the association of NFL, NFM, NFH and α-internexin. In the peripheral nervous system, the neurofilament triplet proteins are associated with peripherin instead of α-internexin (Beaulieu et al., 1999; Yuan et al., 2006).

Most studies on neurofilament assembly and organization have focused on the triplet proteins: NFL, NFM, and NFH, which have molecular weights of 61.5,

102.5, and 112.5 kDa, respectively. Due to the abundance of high negatively charged amino acids along their lengths, and extensive post-transcriptional modifications, their apparent molecular weights rise to 68, 165, and 205 kDa as measured by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-

PAGE).

NFM and NFH, on their own or as a combination, cannot form filaments in the absence of NFL. In contrast, NFL is capable of forming homopolymers in vitro and in human cells in vivo (Geisler and Weber, 1981; Liem and Hutchison, 1982;

Gardner et al., 1984; Hisanaga and Hirokawa, 1988, 1989; Hisanaga et al.,

1990b; Hisanaga and Hirokawa, 1990; Carter et al., 1998; Jacomy et al., 1999).

In rodents NFL, NFM, and NFH are obligate heteropolymers with a 4:2:1

5 stoichiometry (Scott et al., 1985; Lee et al., 1993). This ratio could vary during neuronal development and depending on cellular localization. In the central nervous system, α-internexin is found in a ratio similar to that of NFM in neurofilament polymers (Yuan et al., 2006).

Neurofilament assembly does not require nucleotide binding or hydrolysis but is strongly dependent on ionic strength, pH and temperature (Angelides et al.,

1989). The first step in neurofilament assembly is NFL dimerization with either

NFM or NFH via the association of their conserved rod domains to form parallel side-to-side coiled-coil dimers. Two coiled-coil dimers line up in a half-staggered manner to form an anti-parallel tetramer (Heins et al., 1993). These tetramers combine to form protofilaments, which then associate to constitute the final 10- nm filament (Figure 1.2). Protruding from the filament backbone are the NFM and NFH tail domains (Hirokawa et al., 1984; Hisanaga and Hirokawa, 1988), referred to as neurofilament “sidearms”. The sidearms are important in the stabilization of the cytoskeletal network by forming cross-bridges between neurofilaments and microtubules, and organelles (Figure 1.1).

Phosphorylation of neurofilaments is a recurrent theme in the regulation of filament interactions and organization. Protein kinase A (PKA) and protein kinase C (PKC) can phosphorylate NFL at Serines 51 and 55 (Sihag and Nixon,

1991), and NFM at Serine 23 (Sihag et al., 1999) to regulate in vivo assembly of neurofilaments. NFL phosphorylation at the head domain can prevent assembly

6 and promotes disassembly of pre-existing polymers (Hisanaga et al., 1990a).

The generation of a transgenic mouse with a mutant NFL transgene that mimics a permanently phosphorylated NFL at Ser-55 resulted in pathological accumulation of neurofilaments in brain neuronal cell bodies (Gibb et al., 1998).

In agreement with this result, in vitro polymerization of NFL was inhibited by phosphorylation of NFL head domain by protein kinase N (PKN) (Mukai et al.,

1996).

In 2004, Nguyen et al. discovered that NUDEL was involved in regulating neurofilament assembly. NUDEL is a mammalian homologue of the nuclear distribution molecule NudE present in Aspergillus nidulans. The protein was found to associate directly with the soluble pool of NFL and indirectly with the

NFH subunit. By interacting with NFL, NUDEL promotes the incorporation of neurofilament subunits into the network during neurofilament assembly. Genetic knockdown of NUDEL disrupted the subunit stoichiometry, resulting in impaired neurofilament assembly and transport (Nguyen et al., 2004).

7

Figure 1.2 A model for neurofilament assembly Two neurofilament subunits (NFL and either NFM or NFH) form parallel head-to-tail coiled coil dimers. Two dimers form anti-parallel, half-staggered tetramers. Two tetramers form a protofilaments and four protofilaments come together to form the 10-nm neurofilaments. In total, there are 32 molecules in the cross-sectional area of the filament. Long C-terminal tails of NFM and NFH protrude outwards from the filament (Liu et al., 2004).

1.1.3. Neurofilament function

Expression patterns of neurofilaments are highly correlated with different phases of axonal development. Type III intermediate filament proteins (such as peripherin and vimentin) are present throughout early stages of outgrowth. Once stable synapses have formed at later stages, neurofilaments accumulate robustly as axonal diameter increases, which is also the case during axonal regeneration

(Cochard and Paulin, 1984; Hoffman et al., 1984; Troy et al., 1990; Sánchez et

8 al., 1996). This suggested that an increase in neurofilament number and density correlates directly with an increase in axonal caliber (Friede and Samorajski,

1970).

Axonal caliber and neurofilaments

The first direct evidence implicating neurofilaments in axonal radial growth in an animal model was obtained in Japanese quails. In these quiver quails, neurofilaments are absent from the axons due to a non-sense mutation in the

NFL , leading to neuronal atrophy and a significant decrease in axonal caliber (Yamasaki et al., 1991; Yamasaki et al., 1992; Ohara et al., 1993).

Genetically engineered animals with either enhanced levels of neurofilaments or depleted axonal neurofilaments were used to verify the role of neurofilament density in determining axonal caliber:

In transgenic mice overexpressing human NFH (Cote et al., 1993), perikaryal accumulations of neurofilaments were observed, resulting in a deficiency of axonal neurofilaments and axonal atrophy. The generation of NFH-LacZ transgenic mice in which expression of NFH-β-galactosidase fusion protein provokes the perikaryal aggregation and axonal depletion of neurofilaments caused a 50% reduction of axonal calibers (Eyer and Peterson, 1994; Perrot et al., 2007). Moreover, the targeted disruption of NFL gene in mice resulted in a depletion of axonal neurofilaments and a strong reduction in the diameter of myelinated axons (Zhu et al., 1997).

9

Mice overexpressing NFL showed two- to three-fold increase in the number of neurofilaments but the diameter of their axons was only slightly modified

(Monteiro et al., 1990; Xu et al., 1996), indicating that NFL is not a major contributor to radial growth. Triple heterozygous knockout mice (NFL+/−;

NFM+/−; NFH+/−) in which integrity of the neurofilament network and normal subunit stoichiometry were preserved showed a 40% decrease of neurofilament content and a 50% decrease of axonal diameter in L5 ventral root (Nguyen et al.,

2000). The individual increase in each of the neurofilament subunits inhibited radial axonal growth, and the simultaneous increase of NFM and NFH exacerbated axonal atrophy. In contrast, the co-overexpression of either

NFL/NFM or NFL/NFH increased axonal caliber (Xu et al., 1996; Meier et al.,

1999), suggesting that NFL in combination with either NFM or NFH is sufficient to promote radial growth.

There are two ways for neurofilaments to contribute to axonal caliber: an overall increase in the number of the proteins along the axons, or their space-filling properties, thought to be determined by the long C-terminal tails of NFM and

NFH. To determine the specific contributions of NFM and NFH subunits in axonal size, Elder et al. (Elder et al., 1998a) created NFM-deficient mice. These mice showed a significant reduction in axonal caliber that was accompanied with a decrease in NFL mRNA and protein levels. Modification of NFM expression was also reported in NFL−/− mice (Zhu et al., 1997), suggesting that the levels of

10 NFL and NFM are mutually regulated and that the stoichiometry is important in the regulation of axonal caliber. NFH-null mice, on the other hand, revealed that this subunit contributes to a lesser extent to the determination of axonal diameter

(Elder et al., 1998b; Rao et al., 1998; Zhu et al., 1998), although in these mice the contribution of NFH could have been under-estimated due to a compensatory increase in NFM levels and microtubule density.

It is suggested that the space-filling properties of NFM and NFH C-terminal tails arises from the phosphorylation of KSP repeats on NFM and NFH C-terminal tail domains. Phosphorylation of these domains can regulate axonal caliber through two different ways. Firstly, it can lead to an increase in neurofilament spacing due to the increase in the total negative charges on the side arms causing them to extend laterally by repulsive interactions (Glicksman et al., 1987; Myers et al.,

1987; Chen et al., 2000; Kumar and Hoh, 2004). Secondly, due to increased interfilament interaction mediated by the increased phosphorylation, the axonal transport of neurofilament proteins may slow down (for details refer to section

1.2.5), leading to a local accumulation of neurofilaments, which further contributes to the axonal caliber.

Neurofilaments and conduction velocity

Because radial growth has a direct effect on neuronal conduction velocity, several animal models with abnormal neurofilament expression were utilized to examine the role of neurofilaments in neuronal conduction properties. As

11 mentioned earlier, quiver quails that express a mutant form of NFL suffered from axonal atrophy. This was accompanied by reduced conduction velocity proportional to the decrease in axonal caliber (Sakaguchi et al., 1993). Similarly, lower conduction velocities were observed in NFL−/− mice and NFM−/− mice

(Kriz et al., 2000a), in mice expressing human NFH (Kriz et al., 2000b), in

NFMtailΔ mice (Garcia et al., 2003) and in NFH-LacZ transgenic mice (Zochodne et al., 2004a; Perrot et al., 2007). It should also be noted that, in contrast to

NFH−/− mice, conduction velocity is not altered in NFHtailΔ mice (Garcia et al.,

2003), indicating no implication of NFH sidearm in this parameter.

Mechanical role of neurofilaments

The mechanical importance of neurofilaments as shock absorbers was demonstrated in vitro, where interactions between neurofilaments formed an elastic network, resistant to deformation, a function attributed to the cross- bridges formed by neurofilaments (more details in section 1.1.3; (Leterrier et al.,

1996; Rammensee et al., 2007). Kreplak and his colleagues (Kreplak et al.,

2005) used atomic force microscopy to show that single neurofilaments can be stretched more than three-fold their length without breaking. This suggests that neurofilaments may indeed function as mechanical shock absorbers in vivo.

Perturbations in neurofilament metabolism and organization are associated with various neurodegenerative diseases, including amyotrophic lateral sclerosis

(ALS), Alzheimer ’s disease (AD) and Charcot-Marie-Tooth disease (CMT).

12 1.2. Axonal transport

1.2.1. A historical perspective

The neuron, thought of as the functional unit of the nervous system, is divided into four morphologically and functionally distinct regions: the short dendrites, the neuronal cell body, the long axon, and the axon terminals. Axons can reach lengths exceeding 1 meter in an adult human, and dependent on the cell body for the synthesis of most of their components. Materials destined for the axon are transported anterogradely, toward the axon tip, and materials destined for the cell body are transported retrogradely. This bidirectional transport is known as axonal transport. Although active axonal transport of cargo had been identified decades ago, the history of axonal transport research is relatively short. Weiss and Hiscoe first identified it in 1948 (Weiss and Hiscoe, 1948) in a study performed on regenerating peripheral nerves that were severed in laboratory animals and then were allowed to grow through a restriction in transplanted arteries. They observed a gradual swelling in the proximal axoplasm, due to the accumulation of anterogradely-moving cargo on the site of the constriction.

When the restrictions were removed, the swelling moved along the axons at a rate of ~1 mm/day. Later studies showed that these enlargements were caused by an accumulation of neurofilaments and organelles on the proximal side of the constriction (Schmidt and Plurad, 1985).

This work was followed by studies using pulse injections of radiolabeled amino

13 acids into the cell bodies of dorsal root ganglia neurons by Lasek and his co- workers (Lasek et al., 1984). The amino acids were incorporated into newly synthesized proteins, and transported along the axons. As the wave peak of radiolabeled proteins moved along the axons, the protein composition at any given point varied according to time and distance from the cell body.

Subsequently, two main components of axonal transport were identified: fast transport with velocities ranging between 100-400 mm/day (1-5 µm/second) and slow transport at 0.2-5 mm/day (0.0002-0.05 µm/second) (Tytell et al., 1981).

Fast axonal transport consists of membranous organelles that move rapidly and frequently along axons. On the other hand, the component of axonal transport consists of cytoskeletal and cytosolic proteins like neurofilaments, microtubules and microfilaments.

By that time, two views explaining axonal transport emerged. The first was proposed by Lasek and his colleagues, in which they regarded cytoskeletal and cytosolic proteins as moving along the axon in a slow, synchronous manner, which is now known to be false. The second view, proposed by Sydney Ochs and co-workers, questioned the existence of slow axonal transport and proposed that there is a single fast transport mechanism that encompasses all movement in axons. Slow transport of cargo according to this view is a by-product of fast axonal transport cargos simply dropping off their tracks. The strength of this hypothesis was the recognition that both, fast and slow transport, could be generated by fast movements. However, this hypothesis failed to recognize that

14 slow axonal transport represents a genuine movement of cargoes that are structurally distinct from those of fast axonal transport (Brown, 2008).

1.2.2. Fast axonal transport

At first, the difference in transport rates of the two components was thought to be due to a difference in the mechanism involved in moving cargos along the axons.

Golgi-derived transport vesicles move anterogradely at maximal rates of 200-400 mm/day (2-5 µm/second), while endocytic vesicles, lysosomes and autophagosomes move retrogradely at maximal rates of 100-250 mm/day (1-3

µm/second). The general principles governing fast axonal transport are relatively clear: the organelles move along microtubule and tracks powered by molecular motor proteins (for details refer to section 1.4).

Live cell imaging of membranous organelles along axons revealed that they move in a continuous and unidirectional manner at instantaneous rates that are comparable to the maximal rates of fast transport as determined by radio-isotopic pulse labeling. This means that the cargoes of fast axonal transport move in a highly efficient manner along their tracks. One exception to this notion is mitochondrial transport. These organelles move along the axons with maximal rates of 20-70 mm/day (0.2-0.8 µm/second) (Lorenz and Willard, 1978; Grafstein and Forman, 1980). It is thought that mitochondrial movement is intermittent and bidirectional, making their transport less efficient (Blaker et al., 1981; Morris and

Hollenbeck, 1993; Ligon and Steward, 2000).

15

1.2.3. Slow axonal transport

The slower of the two components of axonal transport can also be resolved into two classes. The slower of the two is referred to as slow component a (SCa), composed of neurofilament proteins, microtubules and microtubule-associated proteins. Slow component b (SCb), on the other hand, is more complex, and includes more that 200 different proteins of which only few have been identified, these include housekeeping proteins like proteins of glycolysis and intermediary metabolism, structural proteins and motor proteins.

Early studies trying to observe slow axonal transport of cargo in live cells assumed a continuous, un-interrupted flow of molecules at the rates observed using radio-isotopic studies. These experiments were performed using fluorescent photobleaching (Lim et al., 1989; Lim et al., 1990; Takeda et al.,

1995) or activation (Okabe and Hirokawa, 1992; Sabry et al., 1995) of labeled cargo. When time-lapse imaging failed to identify the slow, synchronous translocation of fluorescently labeled cytoskeletal polymers, the prevailing view in the field was that these proteins move as unassembled subunits. This became known as the “subunit transport model”.

This model remained the most acceptable explanation for slow axonal transport until two studies in 2000, using cultured neurons and fluorescently labeled proteins, showed neurofilament polymers, and microtubules being transported

16 along neuronal axons using live cell imaging (Roy et al., 2000; Wang et al., 2000;

Shah and Cleveland, 2002; Brown, 2003). The breakthrough was the discovery that cytoskeletal polymers move at fast rates, approaching the rate of movement of membranous organelles, but the average rate of movement is slow because their movements are infrequent, bidirectional, and highly asynchronous.

To observe movement, these researchers took advantage of the discontinuous distribution of neurofilaments along axons of cultured rat sympathetic neurons, which results in naturally occurring gaps in the axonal neurofilament array. Time- lapse imaging revealed the rapid movement of neurofilament polymers through these neurofilament-deficient regions, but the movements were frequently interrupted by prolonged pauses. The neurofilaments moved bidirectionally at peak rates of up to 3 µm/second, which approaches the rate of vesicle transport, but the movements were highly asynchronous and many of the neurofilaments did not move at all during the observation period. These observations made it clear that the actual rate of movement in slow axonal transport is fast, but that the overall rate is slow because the rapid movements are bidirectional, interrupted by prolonged pauses (Brown, 2000). In fact, it was estimated using computational modeling that neurofilaments in motor axons of the mouse ventral root and sciatic nerve spend approximately 97% of their time pausing during their journey along the axon, and similar extents of pausing have been observed experimentally in cultured neurons.

17 1.2.4. Polymers vs monomers

The filamentous appearance of the moving structures in these and other studies indicated that cytoskeletal proteins are transported predominantly as assembled polymers (Roy et al., 2000; Wang et al., 2000; Wang and Brown, 2001; Ackerley et al., 2003; Uchida and Brown, 2004). Yet, not all live imaging studies on GFP- tagged neurofilament proteins supported this view. Some studies have argued that these proteins move predominantly in the form of punctate and short filamentous structures shortly after neuritogenesis (Yabe et al., 1999; Yabe et al.,

2001a; Chan et al., 2003; Helfand et al., 2003b). It was consequently suggested that the punctate structures might represent precursors of neurofilament assembly (Yabe et al., 2001b). However, there was no ultra-structural study of these puncta to differentiate them from short filaments or membrane bound organelles containing fluorescently-labeled proteins. Moreover, the punctate structures could have been an artifact of the high expression levels of exogenous

GFP fusion proteins relative to endogenous proteins.

More evidence supporting the polymer model of transport came in a study by

Yan and Brown (Yan and Brown, 2005). The authors tested the hypothesis that the filamentous structures observed moving through neurofilament gaps along the axons represent single neurofilament polymers using a rapid permeabilization technique to capture moving filaments as they moved through the gaps in the axonal neurofilament array, and observing them using electron microscopy. The result confirmed that the captured filaments were indeed single, continuous 10-

18 nm-diameter neurofilament polymers, and puts to rest the argument that neurofilament polymers do not move at all (Hirokawa et al., 1997; Terada and

Hirokawa, 2000; Terada, 2003).

Despite the heated debate regarding the form in which cytoskeletal proteins are transported along axons, it should be noted that the two perspectives are not necessarily mutually exclusive. In cultured sympathetic neurons, 3% of the moving structures are punctate in shape, and it is possible that these structures represent unassembled neurofilament protein (Uchida and Brown, 2004). In neuronal cell lines, punctate structures predominate at early times after neurite initiation, whereas both punctate and filamentous structures are observed at later stages (Yabe et al., 2001a; Helfand et al., 2003a). Based on these observations,

Yabe et al. (Yabe et al., 1999; Yabe et al., 2001a) have proposed that neurofilament proteins may be capable of movement in both polymerized and unpolymerized forms, that the predominant form varies depending on the differentiation state of the cell, and that the punctate structures observed in these studies represent motile precursors of neurofilament assembly.

1.2.5. Neurofilament phosphorylation and transport

Considerable evidence has pointed to phosphorylation as the most likely mechanism for regulating neurofilament transport and distribution along the axons. Most phosphorylation sites are in KSP motifs of the tail domain of NFM and NFH, where increased phosphorylation correlates with a decrease in the rate

19 of neurofilament transport (Lewis and Nixon, 1988; Watson et al., 1989; Archer et al., 1994; Toyoshima and Komiya, 1995; Jung and Shea, 1999; Yabe et al.,

2001b; Shea et al., 2004).

Hypophosphorylated neurofilament subunits were selectively recovered within a standard microtubule-associated protein preparation rich in kinesin (Saxton,

1994), indicating the association of the neurofilament protein with the tracks and the molecular motor. NFH bearing a developmentally-delayed carboxy-terminal phospho-epitope was selectively not co-precipitated by an anti-kinesin antibody

(Jung et al., 2000). Furthermore, hypophoshorylated NFM and NFH isoforms were transported approximately twice as fast as their extensively phosphorylated counterparts as demonstrated by autoradiographic analysis following metabolic radio-labeling (Jung and Shea, 1999; Jung et al., 2000). Upregulated neurofilament phosphorylation in the sidearm domain has also been associated with slowing of axonal transport of neurofilament protein in cultured neurons in response to glutamate, which can activate members of the mitogen-activated protein (MAP) kinase family, which subsequently phosphorylate the neurofilament side arm domain (Ackerley et al., 2000).

In the optic nerve, it has been clearly shown that signaling from the myelinating oligodendrocytes triggers phosphorylation of KSP repeats in the tail domains of

NFH and, to a lesser extent NFM, which results in local accumulation of neurofilament proteins (Nixon et al., 1994; Sánchez et al., 1996; Sanchez et al.,

20 2000). Axons of retinal ganglion cells are unmyelinated within the retina and for a distance of ~100 µm after they converge and form the optic nerve. Beyond the first 100-150µm, 95% of the axons in the optic nerve are myelinated, large in caliber and contain abundant neurofilaments (Nixon et al., 1994). Under the control of signals from oligodendrocytes, the cross-sectional areas of these axons increase an additional 200% during postnatal development (Sánchez et al., 1996). More than 80% of this radial axonal growth occurs between 21 and 30 days postnatally (Sanchez et al., 2000), and is associated with regional accumulation of neurofilament proteins.

To investigate the relationship of NFH carboxy-terminal phosphorylation to neurofilament organization, the levels of individual NFH phospho-epitopes during axonal radial growth were analyzed and RT97 phospho-epitope expression selectively coincided with the onset and rise of regional neurofilament accumulation (Sanchez et al., 2000). In myelin-deficient shiverer mice, RT97 phospho-epitope levels are selectively reduced and the mice exhibit decreased regional neurofilament accumulation.

To further investigate the role of phosphorylation at the RT97 phospho-epitope in the regulation of neurofilament transport, Miller and his colleagues transfected cortical neurons with GFP-tagged NFH mutant-mimicking permanent phosphorylation or dephosphorylation at this epitope (Ackerley et al., 2003). Live cell imaging studies demonstrated that the mutants mimicking permanent

21 phosphorylation spend twice the time pausing compared to those that mimic dephosphorylation. Consistent with this data, neurofilament transport was accelerated by the application of roscovitine, which is an inhibitor of the kinase

Cdk5/p35 that produces the RT97 phospho-epitope. These observations provide strong evidence for an important role for neurofilament phosphorylation in the regulation of axonal transport.

1.3. Neurofilaments and neurodegenerative disease

The misfolding and aggregation of brain proteins leading to the accumulation of abnormal filamentous deposits in diverse central nervous system cell types are considered early pathological hallmarks of many neurodegenerative diseases.

More than 100 years ago, neurofibrillary tangles and senile plaques were recognized as significant lesions of Alzheimer’s disease, followed shortly thereafter by the discovery of Lewy bodies in the brains of patients with

Parkinson’s disease (Higuchi et al., 2002; Trojanowski and Mattson, 2003;

Forman et al., 2004; Norris et al., 2004). It is yet to be confirmed that these

filamentous protein aggregates actually contribute to the onset and progression of neurodegeneration, are a consequence of other mechanisms that trigger degeneration, or play a neuroprotective role.

Regardless of their contribution to the progression of the disease, intracellular accumulations of proteins can result from one or more of the following

22 pathological processes: (1) abnormal synthesis and folding; (2) aberrant interactions with other proteins; (3) impaired degradation and turn over of the proteins; (4) impaired of disease proteins, especially those targeted for axonal transport over long distances (Roy et al., 2005).

The discovery of mutations in neurofilament proteins associated with several neurodegenerative diseases argues for the participation of neurofilaments in diseases like amyotrophic lateral sclerosis (ALS), Alzheimer’s disease (AD),

Parkinson’s disease (PD), Charcot-Marie-Tooth disease (CMT), giant axonal neuropathy (GAN), dementia with Lewy bodies, spinal muscular atrophy (SMA), progressive supra-nuclear palsy and diabetic neuropathy.

1.3.1. Amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis, also called Lou Gehrig’s disease, is a late-onset progressive motor neuron disease characterized by intraneuronal neurofilament aggregates in affected neurons (Carpenter, 1968; Averback, 1981; Delisle and

Carpenter, 1984; Hirano et al., 1984; Munoz et al., 1988; Murayama et al., 1988).

About 90% of ALS cases are sporadic while approximately 10% are inherited in a dominant manner. Evidence that aberrant neurofilament accumulation can contribute to neuronal death came from the observation that the expression of a mutant NFL subunit causes the aggregation of neurofilaments leading to a selective degeneration of spinal motor neurons and to a severe atrophy of skeletal muscles (Lee et al., 1994).

23

Although the mechanisms leading to the accumulation of neurofilaments in ALS remain unclear, the overexpression of NFL (Xu et al., 1993), NFM (Wong et al.,

1995) or human NFH (hNFH; (Cote et al., 1993) leads to neurofilament aggregations and morphological alterations similar to those found in ALS.

Remarkably, in 1999, Meier et al. were able to rescue the hNFH-induced motor neuron disease by overexpressing human NFL (hNFL). The study reports that the motor neuron disease can be rescued by overexpression of hNFL in a dose- dependent fashion, and that the additional hNFL led to reduction of perikaryal swellings, rescue of axonal transport inhibition and restoration of axonal radial growth.

Mutations in superoxide dismutase-1 (SOD1), the most abundant cytosolic enzyme, account for 20% of all the familial cases among ALS patients. Mice expressing mutant SOD1 display neurofilament accumulations (Tu et al., 1996;

Borchelt et al., 1998) and exhibit a phenotype similar to that of mice overexpressing NFL or hNFH. Interestingly, it was also reported that neurofilament axonal transport is disrupted in mice with SOD1 G37R, G85R, and

G93A mutations (Zhang et al., 1997; Borchelt et al., 1998; Williamson and

Cleveland, 1999). To determine whether neurofilaments are involved in SOD1- mediated disease, mice expressing mutant SOD1 were mated with transgenic mice with altered neurofilament protein content. In a study using NFH-β-

24 galactosidase fusion protein to withhold neurofilaments from the axons of SOD1

G37R mice (Eyer et al., 1998) and in G85R mutants lacking NFL protein

(Williamson et al., 1998), the onset and progression of the disease were significantly slowed and the selectivity for motor neuron toxicity was reduced.

Surprisingly, overexpression of mouse NFL or mouse NFH in SOD1 G93A mice

(Kong and Xu, 2000) and overexpression of hNFH in SOD1 G37R mice

(Couillard-Després et al., 1998) also increase their life span by, respectively,

15% and 65%, which suggests a protective effect of perikaryal accumulation of neurofilaments in motor neuron disease caused by mutant SOD1. However, the mechanism of protection is still unclear

1.3.2. Alzheimer’s disease

Although the involvement of neurofilament protein alterations in the progression of AD is not completely understood, it is known that the perikaryal formation of neurofibrillary tangles (NFT), composed of tau, neurofilaments and other cytoskeletal proteins, is considered an early pathological marker of the disease.

AD is the most common type of dementia characterized by progressive cognitive deterioration and excessive loss of memory, accompanied by declining activities and behavioral changes. The mechanism responsible for NFT formation is not completely understood, but it is reported that tau epitopes are more abundant than neurofilament epitopes in NFT (Schmidt et al., 1990). Neurofilaments in

NFT appear to be more compact, and more extensively phosphorylated than

25 normal neurofilaments (Selkoe et al., 1982; Sternberger et al., 1985; Lee et al.,

1988a).

1.3.3. Parkinson’s disease

PD is a progressive disorder of the CNS affecting dopaminergic neurons of the substantia nigra. Mutations in the parkin gene are the major cause of familial PD

(Leroy et al., 1998). A neuropathological hallmark of PD is the formation of ubiquitinated protein inclusions named Lewy bodies, composed of neurofilaments, α-synuclein, ubiquitin, and proteasome subunits (Goldman et al.,

1983; Galloway et al., 1992; Spillantini et al., 1998). Neurofilaments in Lewy bodies undergo inappropriate phosphorylation and proteolysis (Pappolla, 1986).

A point mutation in a highly conserved region of the rod domain of the NFM gene was reported in a French-Canadian patient who developed the disease at the age of 16 (Lavedan et al., 2002). However, other patients with the same mutation were not affected, arguing against the implication of NFM mutation in pathogenesis of PD. This suggests that mutations in neurofilament are not a primary cause of PD even if the rare variants of the NFM gene identified

(three in total) may act as susceptibility factors.

1.3.4. Charcot-Marie-Tooth disease

CMT is the most common inherited motor and sensory neuron disease (Skre,

26 1978). Patients with CMT progressively develop a weakness of muscles and become unable to walk. CMT neuropathies are classified into several categories, including CMT1, CMT2, CMT3, CMT4 and CMTX. Accumulation of neurofilaments in CMT2 was reported for the first time by Vogel et al. (Vogel et al., 1985). Genetically, CMT is a very heterogeneous group of diseases. CMT2 is autosomal dominant, with various mutations associated with different loci.

Mutations of NEFL gene are associated with CMT type 2E or CMT type 1F. The first mutation was identified by Mersiyanova et al. (Mersiyanova et al., 2000) in a large Russian family with CMT2, while the second was a substitution mutation reported in members of a Belgian family with a severe CMT phenotype (De

Jonghe et al., 2001). Interestingly, the same mutations in cultured neurons disrupted neurofilament assembly and transport and induced mitochondrial accumulations in cell bodies and proximal axons (Brownlees et al., 2002). These mutations also affect anterograde and retrograde fast axonal transport and cause fragmentation of the Golgi apparatus and degeneration of neuritic processes in cultured neurons (Perez-Olle et al., 2002), providing possible mechanisms by which these mutants could be involved in axonal degeneration and CMT pathogenesis.

Examination of nerve biopsies in patients with CMT2 revealed a primary axonopathy characterized by giant axons with swellings composed almost entirely of aggregated neurofilaments (Fabrizi et al., 2004). Finally, an implication of neurofilaments in demyelinating CMT cannot be excluded since

27 nerves from patients expressing NFL mutations show evidence of Schwann cell and neurofilament-phosphorylation abnormalities in demyelinated axons (Fabrizi et al., 2007).

1.3.5. Neuronal intermediate filament inclusion disease

Neuronal intermediate filament inclusion disease (NIFID), also called neurofilament inclusion disease, is a recently described neurological disorder of early onset with a heterogeneous clinical phenotype, including fronto-temporal dementia and pyramidal and extra-pyramidal signs. Symptoms include behavioral and personality changes and, less often, memory loss, cognitive impairment, language deficits and motor weakness (Cairns et al., 2003; Josephs et al., 2003). The pathological phenotype consists of neuronal loss, gliosis, swollen neurons and presence of large neurofilament-rich inclusions in the cell body of neurons (Uchikado et al., 2005). Surprisingly, aggregates are more abundant in areas with little neuronal loss as compared to sites of intense neuronal degeneration. Cairns et al. (Cairns et al., 2003) proposed that the formation of these inclusions is an early event in the pathogenesis of NIFID and that the aggregates are released and degraded into the extracellular space following degeneration of the neurons. The precise mechanism leading to the formation of these aggregates is still unknown.

28 1.3.6. Diabetic neuropathy

Diabetic neuropathy is a peripheral nerve disorder caused by diabetes affecting sensory nerves and dorsal root ganglia and characterized by slow conduction velocity, impairment of axonal transport, axonal atrophy and a reduced capacity for neuronal regeneration. Consistent with this, multiple abnormalities of neurofilament biology have been identified in models of diabetes. Medori et al.

(Medori et al., 1985; Medori et al., 1988) observed in rats with induced diabetes an impairment of the axonal transport of neurofilaments, actin and with a proximal increase and a distal decrease of axonal cross-sectional area accompanied by an important loss of neurofilaments (Yagihashi et al., 1990).

Accumulations of highly phosphorylated neurofilament epitopes are present in proximal axonal segments of dorsal root ganglia sensory neurons from diabetic patients (Schmidt et al., 1997). Moreover, there was a significant decrease in mRNA levels of the three neurofilament subunits as well as reduced neurofilament numbers and densities within large myelinated sensory axons

(Scott et al., 1999). All these results suggest that neurofilament abnormalities may contribute to the development of diabetic neuropathy or may be affected by this disease.

A study by Zochodne et al. (Zochodne et al., 2004b) analyzed the effect on diabetes in transgenic NFH-LacZ mice characterized by neurofilament-deficient axons. They observed an accelerated diabetic neuropathy and increased axonal atrophy in affected mice. This indicates that changes in neurofilament expression, transport or post-translational modifications may be contributing

29 factors to diabetic damage.

1.3.7. Giant axonal neuropathy

GAN is a rare progressive neurodegenerative disorder affecting both PNS and

CNS and generally appears in infancy or early childhood (Berg et al., 1972; Igisu et al., 1975). The disease is characterized by a decline in mental function, loss of control of body movement, and seizures. GAN is caused by mutations in the

GAN gene, which codes for the protein (Bomont et al., 2000). The major cytopathological hallmark is the presence of masses of neurofilaments, producing focal enlargements in the distal regions of axons associated with a reduced number of microtubules. In contrast, axonal segments proximal to the swellings exhibit a reduction in number of neurofilaments (Asbury et al., 1972).

Disorganization and accumulation of other types of intermediate filaments are also found in skin fibroblasts, Schwann cells and muscle fibers (Fois et al., 1985;

Yang et al., 2007). The mechanism of distal axonal accumulation of neurofilaments is still unclear, but an acceleration of their axonal transport was observed in optic nerve from experimentally induced GAN rat model (Monaco et al., 1985).

1.4. Molecular motors

Active transport of molecular cargo along neuronal axons is critical for cellular organization and function. Impairment of this movement could result in neuronal

30 degeneration. Long-distance transport of molecular cargos along neuronal axons takes place along microtubule tracks. Axonal microtubules have a uniform polarity- with their plus ends pointing away from the cell body- and form an overlapping set of tracks that span the entire length of the axon. This orientation determines the direction that microtubule-associated motor proteins carry their cargo.

The main microtubule-based motors are members of the kinesin superfamily, which moves cargo towards the plus end of microtubules, and cytoplasmic dynein, which moves towards the minus end of microtubules. The motors generate force from ATP hydrolysis to move along their tracks. It is increasingly clear that many cargos are moved by both sets of motors, and frequently reverse course. However, microtubules are not the only tracks along which axonal transport takes place. Neurons have multiple compartments, such as dendritic spines and even subcompartments within axons that have few, if any, microtubules. Transport within such regions is thought to be actin-based, and therefore is dependent on myosin motors.

31 Figure 1.3. Principal members of kinesin superfamily proteins Diagrams constructed on the basis of electron microscopy or predicted from primary structure analysis. Large orange ovals correspond to motor domains. Kinesin-1 forms a homodimer with KLC (in blue) associated with the C-terminal tail. Kinesin-3 is monomeric and globular, two isoforms are shown here. Kinesin-13 forms a homodimer with motor domains in the middle. Kinesin-2 is a heterodimer of two isoforms from the same class. Kinesins-4 and -14 are homodimers, but the motor domain in kinesin-14 is at the C-terminus. (Hirokawa, 1998).

1.4.1. Kinesin superfamily

The kinesin superfamily (KIF) is a large gene family of microtubule-dependent motors with 45 members in mice and humans (Aizawa et al., 1992; Miki et al.,

2001). Structurally, they are characterized by a ~360-residue globular domain.

This well-conserved domain contains a catalytic pocket for ATP hydrolysis, and a microtubule-binding site (Hirokawa et al., 1989; Aizawa et al., 1992). Most kinesins form a long filamentous structure, with a globular “head” domain at one end (the site of ATP hydrolysis) and a “tail” domain that associates with light chains at the other end (Figure 1.3). The “stalk/tail” domain is important for the interaction with other subunits of the holoenzyme or with cargo (Diefenbach et al., 1998; Kanai et al., 2004). A short region between the ‘head’ and ‘stalk’, called the ‘neck’, often contains family-specific features. In several families, this

‘neck’ has been shown to be essential for properties such as the direction of

32 motility (Endow and Waligora, 1998) or regulation of activity. The catalytic core and the adjacent “neck”, together, are referred to as the “motor domain”.

Kinesins with a catalytic core at their N-terminus (N-kinesins) have a plus-end- directed motility, while kinesins with the catalytic core at the C-terminus (C- kinesins) have minus-end-directed motility. Some members of the superfamily have their catalytic cores in the middle of their sequence (M-kinesins), and are reported to have a microtubule depolymerizing function.

The “conventional kinesin” (kinesin-1, previously known as KIF5 in mice) was the first member to be discovered. It was identified biochemically as a candidate microtubule-dependent motor for anterograde fast axonal transport (Brady, 1985;

Vale et al., 1985). Subsequently, molecular cloning identified KIF 1-5, which are murine kinesins that are homologous to conventional kinesin at their motor domains (Aizawa et al., 1992). Recently, researchers in the field established a new standardized nomenclature for the classification of kinesins based on 14 large families, kinesin-1 to -14, to facilitate understanding of the evolutionary relatedness of genes that have been identified with different names in various phylogenies (Lawrence et al., 2004). In vitro, kinesins show microtubule- dependent motor activity, and each kinesin has a characteristic velocity. The velocities, ranging from 0.2 µm/sec to 1.5 µm/sec, are consistent with the speed of fast axonal transport in vivo (Hirokawa, 1998b).

33 Kinesins and neurofilament transport

Several studies have indirectly implicated kinesin-1 in neurofilament axonal transport. Kinesin antibody injections into differentiating PC12 cells result in peripherin retention in the cell bodies and depletion of neurofilaments from the neurites (Helfand et al., 2003c). Another study on conditional kinesin-1 knock out mice exhibited neurofilament accumulations in the soma of sensory neurons and reduction in axon caliber (Xia et al., 2003). These studies both suggest a role for kinesin-1 in neurofilament anterograde transport, but neither one monitored neurofilament movement directly. Shea et al. reported phosphorylation- dependent interaction of NFM and NFH with kinesin in situ, and observed neurofilament movement along axons of DRG neurons and neuroblastoma cells in association with kinesin and microtubules. This movement was inhibited using kinesin antibodies (Yabe et al., 1999; Yabe et al., 2000).

In mammalian cells, three genes of kinesin-1 have been identified: kinesin-1A and kinesin-1C are expressed in neurons, while kineins-1C is ubiquitously expressed (Navone et al., 1992; Xia et al., 1998). Studies from our lab using kinesin-1A knock out mice used live-cell fluorescence imaging in cultured SCG neurons to monitor neurofilament transport as compared to wild type controls.

The data show an inhibition of anterograde neurofilament transport in the absence of kinesin-1A. Kinesin-1B and C partially rescued anterograde transport disruption, suggesting some redundancy in the anterograde motor function with respect to neurofilament movement. The data also show an inhibition of

34 retrograde transport, which implies a role for kinesin-1 in the activity of the retrograde motor.

To determine whether this dependence was reciprocal, it was interesting to investigate the role of dynein/dynactin, the retrograde motor, in the transport of neurofilaments.

1.4.2. The dynein/dynactin complex

Dynein is a minus-end directed microtubule-based, motor protein that utilizes energy from ATP hydrolysis to move molecular cargo in cells with 8 nm steps

(Paschal and Vallee, 1987; Hirokawa et al., 1990). Cytoplasmic dynein has been implicated in a variety intracellular motility processes including protein sorting between apical and basolateral surfaces of epithelial cells, the distribution and redistribution of lysosomes, endosomes and Golgi apparatus, ER to Golgi trafficking, chromosomal migration, and retrograde axonal transport. The protein itself is a huge complex of various subunits with a molecular weight of ~2 million

Da. It is composed of two heavy chains (~530 kDa), two intermediate (74 kDa), four light intermediate (53-59 kDa), and several light chains (8-20 kDa) (Figure

1.4).

The two heavy chains of dynein can fold to form globular heads that are comprised of multiple AAA-like domains (AAA1 -> AAA6) that form a heptad ring structure (Ogura et al., 2004). Affinity of the dynein motor to microtubules is

35 regulated by ATP binding and hydrolysis at AAA1 site, although AAA3 is also thought to bind ATP (Ross et al., 2006). The microtubule binding sites are on two flexible stalks that project from the head domain between AAA4 and AAA5

(Gee et al., 1997).

The basal domain of dynein includes the remaining subunits, and is thought to be the site of cargo binding. The dynein intermediate chains (DIC) are important for binding cargo and dynactin (Karki and Holzbaur, 1995; Vaughan and Vallee,

1995; Steffen et al., 1996). Vaughan and Vallee have identified two DIC genes and five splice variants in rat brain (Vaughan and Vallee, 1995).

In vivo, the activity of dynein requires dynactin, which is thought to increases the processivity of the motor by binding to microtubule tracks as well as cargo (King and Schroer, 2000b). It is formed of 11 subunits that can be divided into two distinct structural domains: an actin-like mini-filament backbone which includes dynamitin (also known as p50), and a flexible projecting sidearm formed by p150Glued and p24 (Figure 1.4; (Schafer et al., 1994; Allan, 1996; Schroer et al.,

1996; Eckley et al., 1999).

p150Glued is the largest subunit in the dynactin complex and acts as a binding partner of both microtubules and dynein (Karki and Holzbaur, 1995; Vaughan and Vallee, 1995; Waterman-Storer et al., 1995; Quintyne et al., 1999; Vaughan et al., 2001). The microtubule binding sites are on the distal end of p150Glued,

36 and are important in increasing dynein/dynactin processivity by attaching the complex to the microtubule tracks even when the motor itself is disengaged.

Overexpression of full-length p150Glued disrupts dynein-based motility, as assessed by immunolocalization of centrosome and endomembrane components

(Quintyne et al., 1999). Moreover, p150Glued can be released from dynactin by chaotropic salts or an excess of dynamitin/p50 (Echeverri et al., 1996; Karki et al., 1998; Eckley et al., 1999). Therefore, upon dynamitin overexpression, dynein can still bind the shoulder/sidearm, but lacks a mechanism for binding cargo, which leads to a wide variety of motility defects (Burkhardt et al., 1997;

Waterman-Storer et al., 1997; Melkonian et al., 2007).

Figure 1.4. The dynein/dynactin complex A schematic representation of the major subunits forming cytoplasmic dynein: Dynein heavy chain, intermediate chain and the light intermediate chain. They are shown here in association with the subunits forming dynactin: p150Glued, p50 (dynamitin) and Apr1. (Hirokawa, 1998a).

37 Dynein/dynactin and neurofilament transport

There are several studies that suggest a role for dynein/dynactin in the transport of neurofilaments. Neurofilaments were purified in their native purified state from mammalian spinal cords and visualized moving bidirectionally on microtubule tracks in vitro. This movement was disrupted upon treatment with a function blocking DIC antibody (74.1), suggesting a role for dynein in neurofilament transport. Biochemical tests showed that the neurofilament sample co-purified with DIC, DHC, p150Glued, and p50/dynamitin (Shah et al., 2000). Wagner et al.

(Wagner et al., 2004) later discovered that the interaction between dynein and neurofilaments is the result of DIC and NFM interaction. In PC12 cells dynein was found to co-localize with peripherin. Overexpression of dynamitin/p50 in these cells caused an accumulation of neurofilaments at the distal end of neurites, suggesting a disruption of retrograde transport (Helfand et al., 2003c).

A role for dynein in neurofilament transport was also reported in a study of microtubule transport by He et al. (He et al., 2005). The authors used live-cell fluorescence imaging of rat SCG neurons transfected with DHC siRNA to monitor the frequency of neurofilament axonal transport. They observed a dramatic decrease in the frequency of retrograde transport (92%), suggesting that dynein is the retrograde motor for neurofilaments. Anterograde movements, on the other hand, increased in frequency (64%), indicating that inhibition of the retrograde motor enhances anterograde transport. This result does not agree with the result obtained from our kinesin experiments that suggested a functional

38 interdependence between kinesin and dynein. On the contrary, the result from

He et al. indicates that dynein and kinesin are independent.

On the basis of these observations, and to investigate the regulation of bidirectional axonal transport of neurofilaments, we set out to investigate the effect of dynein/dynactin disruption on the retrograde and anterograde transport of neurofilaments in neuronal cells in culture.

Figure 1.5. Schematic representation of three classes of myosin The basic structure of I, V and VI which are involved in neuronal transport. Myosin I lacks a coiled coil region and does not form dimers, in contrast to myosins V and VI. (Bridgman, 2004).

39 1.4.3. The myosin superfamily

Members of the myosin superfamily are extremely diverse actin-binding motor proteins that are characterized by three main domains: an N-terminal motor head domain that binds actin and ATP, a neck domain consisting of one or more light chain binding IQ motifs, and a C-terminal tail domain (Figure 1.5; (Cheney et al.,

1993; Berg et al., 2001; O'Connell et al., 2007). Genomic studies are continuously identifying new classes of myosin: a recent study using comparative genome analysis revealed a total of 37 different motor domains (Richards and

Cavalier-Smith, 2005). Members of the myosin superfamily share a conserved

~80 KDa domain on their N-terminus. The C-terminal tail is considered class- specific, determining properties such as membrane binding or kinase activity

(Krendel and Mooseker, 2005). Some of the differences between myosin classes are listed in Table 1.1.

Neuronal myosins include classes I, II, V, VI, and IX. The first four classes have been implicated in transport in neurons and other cell types. Unlike myosin I, myosin II and myosin V both possess an α-helical domain that allows dimerization. These three myosins also differ in the type and number of light chains present in the neck region. The light chains in myosin I and myosin V are calmodulin, a calcium-binding regulatory protein in many intracellular enzymes.

Like other molecular motors, the head domain hydrolyses ATP to generate mechanical energy. This process is actin-activated. The tail domains of myosin I and myosin V can bind the plasma membrane and the membranous organelles,

40 which affects their cellular functions. Myosin II rod-like tail domains, on the other hand, associate with each other to form thick filaments, which form part of the contractile machinery in muscle, allowing multiple myosin heads to interact with the actin filaments simultaneously.

Myosin Number of heads Processivity Duty ratio Directionality I 1 No Low + IIa 2 No Low + IIb 2 n.a. High + III 1 No n.a. + V 2 Yes Moderate + VI 1 Yes High - VII 1 or 2 Yes High + IX 1 Yes n.a. + X 2 n.a. Moderate - XI 2 Yes High +

Table 1.1. Structural and mechanical characteristics of various myosin classes Different classes of myosin have different measures of processivity that are mainly dependent on their ability to dimerize. Class V myosins are especially fast motors that average ~300/400 nm/second and move towards the plus-end of actin filaments. (Krendel and Mooseker, 2005).

Myosin V

Myosin V is involved in fast axonal transport of vesicular cargo in neurons

(Langford and Molyneaux, 1998; Mermall et al., 1998; DePina and Langford,

1999). Myosin V dimerizes to form a processive, two-headed motor protein

(Figure 1.5) that moves towards the barbed (plus)-end of the filaments in vitro.

Three forms of myosin V that are differentially expressed within tissues and organs have been identified. Myosin Va, for example, is highly abundant in the

41 brain and nervous tissue (Mercer et al., 1991). Myosin Vb is also present in the brain, but with less abundance, and is expressed in other tissues as well (Zhao et al., 1996). The third form, myosin Vc, is present in low levels in the cerebellum, and is highly abundant in epithelial cells (Rodriguez and Cheney, 2002). In neuronal cells, myosin V was present in organelle-rich regions of the growth cones, actin filaments and the plasma membrane (Evans et al., 1997).

Immunofluorescence and immunoelectron microscopy revealed myosin V- associated organelles on microtubules and actin filaments. SCG extracts taken from wild type and dilute lethal mice that lack myosin Va were used to study the effect of myosin Va on neurite outgrowth. Neurons lacking the motor were capable of extending neurites with normal lengths and well-spread growth cones with filopodia and lamellipodia that were not different from wild type (Evans et al.,

1997).

The role of myosin Va in axonal transport has been more closely studied than for the other myosins. It is one of the fastest non-muscle myosins identified, moving at rates 300-400 nm/second (Cheney et al., 1993). Previous studies have identified a role for myosin Va in transport in the melanocytes, where the motor binds to melanosomes, even as they are transported along the microtubule tracks. As these melanosomes reach the peripheral segments of microtubules, myosin Va can transport them into the actin meshwork of the dendrites.

Disruption of myosin Va activity results in perinuclear accumulation of melanosomes, leading to loss of pigmentation from hair and skin (Wu et al.,

42 1998). Myosin Va was also identified on the surface of membranous organelles moving along microtubule tracks in neurons (Bridgman, 1999). In the absence of the motor, organelles associated with synaptic vesicle proteins accumulate in the peripheral sites that include pre-synaptic terminals.

The ability of myosin Va to transfer molecular cargo from one cytoskeletal track to another is supported by its association with rough endoplasmic reticula (RER), possibly contributing to the distribution of these organelles in the cells, but also facilitating the transfer of organelles from their site of formation to the microtubule-dependent fast transport tracks that enter axons and dendrites

(Bridgman, 2004).

Figure 1.6. Increased neurofilament content and density in dilute lethal axons Electron micrograph images show neurofilament numbers (insets) are increased in the axoplasm of sciatic nerve neurons from dilute lethal (electron micrograph on far left) mice as compared to wild type although there was no change in the general organization of the axoplasm. The numbers from 160 axons from wild type (black boxes) and dilute lethal (red boxes) are plotted against axonal cross sectional areas revealing a significant increase in neurofilament density in mice lacking myosin Va. (Rao et al., 2002).

43

Myosin Va and neurofilament transport

The interaction between myosin Va and neurofilaments in the neuronal axons was investigated in a study by Rao et al. (2002). Ultrastructural observations of immunogold-labeled sections of mouse optic nerve axons revealed that myosin

Va is associated with neurofilaments. Using blot overlay analysis, the authors analyzed the association of myosin Va with the different neurofilament subunits:

NFL, -M, and -H, and reported that of the three subunits, only NFL was bound to myosin Va. These results show that myosin Va is capable of binding neurofilaments, specifically NFL, in vivo. Based on this, the authors tested a possible relationship between NFL content and myosin Va in mice that were lacking the NFL gene. In such mice, the myosin Va content was reduced by 55% in the sciatic nerve axons. In mice overexpressing NFL, there was a 1.5 fold increase in myosin Va, suggesting an interaction between the two proteins. The authors also used dilute lethal mice from the DLS/LeJ strain that lack functional myosin Va protein to assess the physiological significance of the myosin Va- neurofilament interaction. Immunoelectron microscopy showed that the neurofilament number was increased nearly two-fold in the peripheral axons of dilute lethal mice, but this was not accompanied by an increase in the cross- sectional area of the axons (Figure 1.6). Such accumulations could be explained by a disruption of neurofilament movement along the axons, but this was not directly tested in the study. To test this hypothesis, we investigated the axonal

44 transport of neurofilaments in cultured neurons from SCG and DRG of wild type and dilute lethal mice using live-cell imaging.

45

Chapter 2: Materials and Methods

2.1. Mice

2.1.1. ICR mice

Wild type ICR pregnant female mice were obtained from Harlan Laboratories

(Indianapolis, IN) and used for our dynein/dynactin study. For our cortical neuronal cultures, the females were sacrificed and the embryos were used at

E16.5 (as described in section 2.2.4). For SCG and DRG neuronal cultures, post-natal pups were used (as described in sections 2.2.1 and 2.2.2).

2.1.2. DLS/LeJ mice

DLS/LeJ mice (JAX GEMM stock no. 000253) were purchased from The Jackson

Laboratory (Bar Harbor, ME). Like all mice homozygous for the dilute lethal spontaneous mutation, DLS/LeJ mice display a severe neuromuscular disorder characterized by convulsions and opisthotonus (a case of hyperextension and spasticity in which the head, neck, and spinal column arch backwards) apparent at 8-10 days of age, and they lack smooth endoplasmic reticulum in the dendritic spine of Purkinje cells, causing an absence of intracellular calcium. Loss of

46 these calcium pools is thought to cause the neurological symptoms. These mice usually die by approximately 3 weeks of age.

The dilute-lethal (dl) mutation in this strain of mice is maintained in repulsion with short ear mutations (Bmp5se), which code for bone morphogenic protein 5, which means that the resulting strain is segregating for dl, Bmp5se, and their wild-type alleles. Both mutations are recessive and are closely linked on 5, rendering the probability of mice being homozygous for both mutations at one time highly unlikely.

To identify homozygous dilute mice from this strain, we used the dilute phenotype. Mice homozygous for the mutation showed an apparent lightening in skin color, due to the absence of pigment, which was apparent by day 4 after birth. We confirmed the loss of myosin Va in those mice using Western blot analysis. Control mice in this set of experiments were P4 pups from C57BL/6J mice purchased from Harlan Sprague Dawley (Indianapolis, IN). The DLS/LeJ strain was maintained by inbreeding heterozygous progeny, which were identified by their darker coat color as compared to homozygous dl mutants. The wild-type pups were identified by their dark coat color and short-ear phenotype that arises due to the homozygous Bmpse mutation, which was apparent by 10 days of age.

47 Coat color Ear length

wild type Black Short

dl/dl Dilute (light) Long

dl/wt Black Long

Table 2.1. Phenotypic identification of DLS/LeJ strain The recessive dilute and short ear mutations are maintained in repulsion. As a result, mice wild type for dilute are black in color, but possess a short ear phenotype. Mice homologous for the dilute mutation appear lighter in color, but have normal ear length. Mice heterozygous for the mutations have a black coat color and normal ear length.

2.1.3. dl20J mice

We received the dilute viral strain of mice from Dr. Nancy Jenkins at The Institute for Molecular and Cell Biology in Singapore. These mice are maintained on a

C57BL/6 background and are heterozygous for both dilute lethal and dilute viral mutations. To eliminate the dilute viral mutation, we crossed the progeny with wild-type C57BL/6 mice, and sacrificed all mice carrying a dilute viral allele. The primers we used for genotyping were: dvLTR-F: 5’-CCC GTG TAT CCA ATA AAG CC-3’ dilute-R: 5’-TCC TCT GTG GTC ATC ACT GG-3’ dilute-F: 5’-TGG AAT CCC AGC AGT GGT A-3’ using a protocol provided by Wolfgang Wagner (Laboratory of Cell Biology,

National Institutes of Health/National Heart, Lung, and Blood Institute).

As for the DLS/LeJ strain, the recessive dl mutation in these mice is also maintained in repulsion with the Bmp5se mutation on chromosome 9. To get rid

48 of this short ear mutation, mice were crossed with wild-type C57BL/6 mice. The heterozygous mice were then inter-bred to maintain the colony. To identify the mice, tail snips were collected at the time of weaning (~21 days), and genotyped using four primers: dl20J-1: 5’-CAC CAT CAT CTC ATT TCC ATC CTG TGT CC-3’ dl20J-2: 5’-CTC AGG AGG ATA ATA AAT GCA CGA GAC GC-3’ dl20J-3: 5’-CTC ATC TAT ACA TGG TAA TAG CAG GTG GC-3’ dl20J-4: 5’-CAG TTA GAG AAG GCT AGA AGT AGC AGA GG-3 using a protocol provided by W. Wagner. All experiments were performed on mice homozygous for the dilute lethal allele. For these experiments, we used wild-type littermates as control.

Figure 2.1. PCR genotyping of dl20J mice Wild type mice were identified by the appearance of two bands (251 and 291 bp). Mice homozygous for the dl mutation were identified by the appearance of 1 band (334 bp). Heterozygotes were identified by the appearance of three bands at 251, 291, and 334 bp.

49 2.2. Cell culture

2.2.1. SCG neuronal culture

Superior cervical ganglia were dissected from P4 mice, dissociated, and cultured at a concentration of 0.2 ganglia/ml on No. 1.5 square glass coverslips (22 X 22 mm) coated with 1 mg/ml poly-D-lysine (molecular weight, 70,000 -150,000;

Sigma, St. Louis, MO) and 10 µg/ml MatrigelTM (BD Biosciences, San Jose, CA) as previously described (Brown, 2003). The cultures were maintained at 37°C in

Leibovitz’s L-15 medium (phenol red free; Invitrogen, Carlsbad, CA) supplemented with 0.6% glucose, 2 mM L-glutamine, 50 ng/ml 2.5S nerve growth factor (BD Biosciences), 10% adult rat serum (Harlan Sprague Dawley), and

0.5% hydroxypropylmethylcellulose (MethocelTM; Dow Corning, Midland, MI). If needed, the medium was changed every three days.

2.2.2. DRG neuronal culture

Dorsal root ganglia were extracted from P3-P7 mouse pups, dissociated and cultured at a concentration of 1 ganglion/ml on 40 mm circular glass coverslips

(Bioptechs, Butler, PA) coated with 1 mg/ml poly-D-lysine (molecular weight,

70,000 -150,000; Sigma) and 10 µg/ml MatrigelTM (BD Biosciences). Cultures were maintained at 37°C in Leibovitz’s L-15 medium (phenol red free; Invitrogen) supplemented with 0.6% glucose, 2 mM L-glutamine, 50 ng/ml 2.5S nerve growth factor (BD Biosciences), 10% adult rat serum (Harlan Sprague Dawley), and

0.5% hydroxypropylmethylcellulose (MethocelTM; Dow Corning, Midland, MI).

50

2.2.3. Astroglial cell culture

Astroglial feeder cultures were prepared as described by the protocol provided by

Barbara Smoody in the lab of Dr. Gary Banker (detailed in Banker and Goslin,

1991). Three P1 ICR mouse pups (Harlan) are decapitated, and their brains removed and placed in a dish containing ice-cold HBSS. The cerebral hemispheres are dissected out, cleaned of the meninges, and placed in an enzyme solution containing trypsin (0.25% w/v) and Dnase (1% v/v) in warm

HBSS. The cells are resuspended in Glial MEM (MEM media containing 10%

[v/v] horse serum, 1.5% [v/v] glucose [45%] and 12 µg/ml gentamicin. The yield is usually 8 to 10 million cells/brain. Cells are diluted to the desired concentration in Glial MEM and plated onto T75 flasks at a density of 400,000-500,000 cells per ml (15 ml total). If the cells are needed for an experiment directly, they are seeded onto 60 mm dishes at a density of 200,000-300,000 cells per ml. Cells are fed the day after plating and twice a week thereafter.

2.2.4. Cortical neuronal culture

Cortical sandwich neuronal cultures were prepared based on the protocol described by Barbara Smoody at Oregon State University (for details, see

Banker and Goslin, 1991). Briefly, brains from E16.5 ICR mouse pups (Harlan) were removed and placed in ice-cold HBSS. The cortices were dissected out, de-sheathed of the meninges, and dissociated using an enzyme solution

51 containing 0.025% Trypsin, 0.01% EDTA and 0.5mg/ml Dnase I (250 ml in 5ml) in PBS. The cells were then plated on 40 mm circular glass coverslips

(Bioptechs) dotted with sterile paraffin and coated with 1 mg/ml poly-D-lysine

(molecular weight, 70,000 -150,000; Sigma) and 1 µg/ml Laminin (BD

Biosciences) at a concentration of ~50,000 cells/ml in Neurobasal medium

(Invitrogen) containing 2% B27 supplement (Gibco), 2 mM L-glutamine, 0.3% glucose, 3.75 mM NaCl, 5% fetal bovine serum (HyClone, Thermo Fisher

Scientific, Waltham, MA), 20 µg/ml gentamicin (Gibco), and 2.5 µM AraC

(Sigma). Two hours after plating, the coverslips were transferred to 60 mm petri dishes containing the glial feeder layer with the paraffin dots facing down to separate the neurons from the glial cells. Three days after plating, the “plating media” is removed and replaced by “culture media” containing the same components except the serum. Culture media is then changed every four days.

2.3. Cloning and transfection

2.3.1. DNA constructs

Mouse NFM cDNA (Genbank accession number DQ201636) was obtained by

RT-PCR using RNA from wild-type P24 mouse cerebellum and sub-cloned into the pEGFP-C1 mammalian expression vector (Clontech, Mountain View, CA).

The resulting pEGFP-mNFM expression vector coded for the codon-optimized

F64L/S65T variant of green fluorescent protein (GFP), fused to the N-terminus of mouse NFM by a 25 amino acid linker (Yan et al., 2007).

52

The photoactivatable construct was created by subcloning the mouse NFM cDNA into the pPAGFP-C1 vector of Patterson and Lippincott-Schwartz (Patterson and

Lippincott-Schwartz, 2002). The resulting pPAGFP-NFM expression vector was identical to the pEGFP-NFM vector in all respects, except for the PAGFP

(photoactivatable GFP) coding sequence.

The pDsRed2 construct was purchased from Clontech. pmCherry was obtained by subcloning mCherry (Shaner et al., 2004) into pEGFP-C1 (Clontech) in place of the EGFP sequence.

The myc-tagged p50/dynamitin was obtained from Dr. Richard Vallee (Columbia

University, New York). Dynamitin was cloned into a Not1 site of the pCMVbeta vector (Clontech). The myc epitope tag (MEQKLISEED-stop) (Evan et al., 1985) was inserted after the last p50 codon by PCR mutagenesis and subcloning through a shuttle vector (Echeverri et al., 1996).

Dr. Peter Baas (Drexel University, Philadelphia, PA) sent the plasmid pGFP-p50 in E. coli JM109, from Dr. Trina Schroer’s lab (Johns Hopkins University,

Baltimore, MD). The construct was derived from chick p50 cDNA (Accession number: AF200744). A fragment of the plasmid (pCDNA3; Invitrogen) containing the pGFP and myc-tag was replaced by pEGFP from pEGFP-C1 (Clontech) to enhance fluorescence.

53

We received the DsRed-CC1 construct from Dr. Nick Quintyne (Florida Atlantic

University, FL). Cloning of the construct was described in Quintyne and Schroer,

2002. Briefly, a fragment containing p150217–548 (another name for p150-CC1) was amplified from CMV-p150 (Quintyne et al., 1999) by PCR, inserted directly into the pTA vector, and then subcloned into pDsRed-N1 (Clontech Laboratories

Inc.) using EcoRI. All plasmids were purified using EndoFree Maxi plasmid purification kits (Qiagen).

2.3.2. RNA interference siRNA oligonucleotide duplexes targeting non-overlapping regions of mouse dynein heavy chain mRNA were purchased from Invitrogen (Stealth™ Select 3

RNAi). The sequences of the triplet were as follows:

(1) 5’-GAG GCU UCG GCA GUA UGC UUC CUA U-3’ and 5’-AUA GGA AGC

AUA CUG CCG AAG CCU C-3’

(2) 5’-CCG GCG UUU CCA GCA UCA UCU UAA A-3’ and 5’-UUU AAG AUG

AUG CUG GAA ACG CCG G-3’

(3) 5’-GAG UGU GCU UGU AAG UGC AGG CAA U-3’ and 5’-AUU GCC UGC

ACU UAC AAG CAC ACU C-3’

For control experiments, we used the corresponding Stealth™ RNAi Negative

Control Duplex with medium GC content (Invitrogen). This duplex is designed to minimize to any known vertebrate transcript, does not induce the interferon-mediated stress response pathways, and demonstrates

54 minimal knockdown of vertebrate target genes.

2.3.3. DNA and RNA injections

Prior to microinjection, we add a thin layer of sterile, warm dimethylpolysiloxane

(5 centistokes, Sigma) over the culture medium to prevent evaporation.

For nuclear injections, the DNA was diluted to 20�mg/ml in 50 mM potassium glutamate, pH 7, and 1.25 mg/ml tetramethylrhodamine dextran (MW 10,000;

Sigma-Aldrich) to permit the visual confirmation of our injections. Transfection was performed by nuclear injection for DNA constructs and cytoplasmic injections for siRNA using an Eppendorf FemtoJet with InjectMan three-axis motorized micromanipulator (Brinkmann Instruments, Westbury, NY).

Micropipettes used for the injection were pulled from thick-wall borosilicate glass

(1.0 mm O.D., 0.58 mm I.D.; World Precision Instruments, Sarasota, FL) using a

Sutter P-97 Flaming-Brown micropipette puller (Sutter Instrument Company,

Navato, CA). Nuclear injections were performed two days after plating (Brown,

2003).

For RNA interference experiments, the lyophilized oligonucleotide duplexes were reconstituted to 20µM in DEPC-treated water and then diluted to 100 nM in 50 mM potassium glutamate, pH 7.2. The oligos were microinjected into the cytoplasm of neurons after one day in culture.

55

2.3.4. Antibody injections

The mouse monoclonal antibody IC74.1 was obtained from Dr. Kevin Pfister

(University of Virginia, VA; (Dillman and Pfister, 1994). Control experiments were performed using Jackson ImmunoResearch Laboratories Chrompure mouse IgG, whole molecule, dissolved in 0.01M sodium phosphate, 0.5M NaCl, pH 8.0.

To prepare the antibodies for microinjection, both antibodies were dialyzed against 50 mM potassium glutamate, pH 7.2, and then concentrated to 2.4 mg/ml in a Microcon centrifugal concentrator (Millipore, Billerica, MA). Five days after plating the cells (i.e. on the day of observation), cytoplasmic injections were performed, with a mixture of 2.4 mg/ml antibody and 1.5 mg/ml tetramethyl- rhodamine-labeled dextran (Mw 10,000, Sigma). Movies were acquired 2-6 hours after injection.

2.3.5. Electroporation

For cortical neuron transfections, suspensions of cells from one brain per transfection were centrifuged at 100 Xg for 5 min at room temperature and resuspended in 100 µl of mouse neuron nucleofection solution (Amaxa

Biosystems, Gaithersburg, MD). The cells were co-transfected with 2.5- or 7.5-

µg of DsRed-CC1 and 2.5 mg of pEGFP-mNFM by electroporation using program O-005 of an Amaxa Nucleofector (Amaxa Biosystems), diluted in 2ml plating media and left to recover at 37°C for 10 min. The cells were then plated

56 at a concentration of ~50,000 cells per 40 mm round coverslip (No. 1.5 glass coverslips; Bioptechs, Butler, PA), as detailed earlier.

2.4. Live cell imaging

2.4.1. Heated chamber system

In experiments involving live-cell imaging of SCG neurons, the microscope stage and objective were pre-warmed using an air stream incubator (Nevtek,

Williamsville, VA; and Nicholson Precision Instruments, Bethesda, MD). The culture dishes were then placed on the stage with the air stream still running to maintain a temperature of 37°C.

For live-cell imaging from cortical and DRG neuronal cultures, the coverslips were mounted in a Bioptechs FCS2 closed-bath heated imaging chamber

(Bioptechs) and maintained on the stage at 37°C for as long as 6 hours in

Hibernate A low-fluorescence medium (BrainBits, Springfield, IL) supplemented with 0.3% glucose, 1 mM L-glutamine, 2% (v/v) B27 supplement, 62.5 mM NaCl,

2 µg/ml gentamicin, and 50ng/ml 2.5S nerve growth factor. The medium was changed every 2 hours using the perfusion ports of the heating chamber. The objective was maintained at 37°C throughout the duration of the imaging using a

Bioptechs objective heater.

57 2.4.2. Neurofilament movement through gaps

The neurofilament population in SCG neurons is characterized by a discontinuous array that gives way to the appearance of naturally-occurring gaps along the length of the axons. The appearance of gaps seems to be directly proportional to the density of neurofilaments along axons, with the highest frequency usually observed along the thinnest axons. These gaps can be utilized to observe the movement of fluorescently tagged neurofilament polymers as they are transported along the axons.

For our observations, we used epifluorescence and either differential interference contrast (DIC) or phase contrast microscopy on a Nikon TE2000 or TE300 inverted microscope (Nikon, Melville, NY) using Nikon 100X/1.4NA Plan Apo DIC or 100X/1.4NA Plan Apo Ph oil-immersion objectives and FITC/EGFP,

TRITC/DsRed, or Texas Red/mCherry filter sets (Chroma Technology Corp.,

Rockingham, VT).

For time-lapse imaging of the neurofilament transport along the axons, we used the neutral density filters to attenuate the fluorescent intensity by 8-12- fold.

Images were taken with 700-1000 msecond exposures with 4-second intervals.

Neurofilament movement was analyzed by tracking the position of the leading end of the filament using the ‘TrackPoints’ function in the ‘Motion Analysis’ module of the MetaMorph software. We tracked the movement of all objects that exceeded 1.3 µm in length and that moved >2.6 µm along the axon in either

58 direction. Neurofilament length was determined using the ‘Region

Measurements’ function in MetaMorph.

Frequency of movement was calculated by dividing the number of moving objects by the total time in minutes. Filaments were classified as either anterograde or retrograde based on their net direction of movement during the tracking period. A filament was considered to reverse direction if it moved for at least 8 µm in the opposite direction. For each time interval, we calculated an interval velocity, which is the distance moved in one time interval divided by the duration of that time interval. The average velocity, including pauses for each filament, was calculated by averaging all the interval velocities for that filament.

The average velocity excluding pauses was calculated by averaging all the interval velocities, except those <1 pixel/second (0.13 µm/s), which we estimate to be the precision limit of our measurements (Wang and Brown, 2001).

Displacements of <0.13 µm/s were considered to be pauses.

Anterograde flux were calculated in each axon by summing up the anterograde movements of all filaments observed moving throughout the duration of the movie, regardless of the net directionality of the filament or the distance travelled, divided by the total imaging time. Similarly, we calculated retrograde flux in every axon by adding the total distances moved in the retrograde direction of all filaments observed regardless of their directionality, and divided the sum by the observation time.

59

2.4.3. Fluorescence activation pulse-escape

For pulse-escape experiments, DRG neurons were cotransfected with pPAGFP-

NFM and either pDsRed2 or pmCherry 2 days after plating. For neurons extracted from DLS/LeJ mice, we used DsRed2 as a fluorescent marker, and we acquired movies 5-7 days after plating. In neurons extracted from the dl20J mice, we used mCherry and acquired movies 5-8 d after plating. Transfected cells were identified based on the presence of DsRed2 or mCherry fluorescence, and a ~20 µm segment of axon was photoactivated by exposing it to violet light for 1 second in the absence of neutral density filters. The region to be photoactivated was defined using a field-limiting diaphragm (Trivedi et al. 2007).

Time-lapse images of the photoactivated green fluorescence were acquired at 5- minute intervals using 1-second exposures for 115 min. To avoid photobleaching, all focusing was performed on the DsRed2/ mCherry channel, and the exciting light on the GFP channel was attenuated 32-fold using neutral density filters. Measurements of the fluorescence intensity in the activated region were performed using MetaMorph software, and the resulting values were corrected for background and expressed as arbitrary analog-to-digital units per micrometer of axon (ADU/µm). When using DsRed2, we also corrected for cross-excitation on the GFP channel as described previously (Trivedi et al.,

2007).

60 If the first frame of the time-lapse series was out of focus, the movie was not analyzed. For subsequent frames that were out of focus, we estimated the fluorescence intensity by linear interpolation between the adjacent time intervals in the series. If more than two consecutive frames were out of focus, the movie was not analyzed.

2.4.4. Calculation of mean time to depart

The mean time of departure is the average time it takes a fluorescent neurofilament to leave the activated region. The fluorescence Intensity as a function of time is proportional to the number of neurofilaments in the activated region along the axon (F(t)). Since it takes only seconds for a neurofilament to move out of the activated region, the velocity of movement has negligible influence on the pulse-escape kinetics, at least on a time course of minutes or hours. Thus, F(t) is determined by the pause durations. If the fluorescence intensity is normalized (i.e.: F(0)=1), then the fluorescence at any point in time can be considered to represent the probability that a neurofilament is still in the activated region at time t. Thus, the probability that a neurofilament leaves the activated region within time interval [t: t+ Δt] is given by: f(t)-f(t+Δt)= (-df/dt)Δt for small time intervals Δt, and the probability density of escape times ρ(t) is given by -df/dt.

It follows that the probability of a neurofilament departing from the activated region in time interval [t: t+Δt] is given by ρ(t)Δt.

61

We have shown previously that the kinetics of departure in the pulse-escape experiments obey a double exponential relationship (Trivedi et al., 2007). Thus the non-normalized fluorescence intensity in the activated region is given by

F(t)=a1exp(-λ1t)+a2exp(-λ2t) in which a1 and a2 are the fluorescence intensities at t=0 attributable to neurofilaments in the mobile (“on track”) and stationary (“off track”) states, respectively, and λ1 and λ2 are the corresponding exponential decay constants.

To normalize to the starting fluorescence intensity, we have to divide by a1 + a2, arriving at the following:

a1 a2

F(t)= ------exp(-λ1t) + ------exp(-λ2t). a1+ a2 a1 + a2

Thus, the probability density of escape times becomes:

a1 a2

ρ(t)= ------λ1exp(-λ1t) + ------λ2exp(-λ2t). a1+ a2 a1 + a2

The mean time to depart (T) is the first moment of the probability density of escape times:

∞

(T) ∫tρ(t) dt= ([a1/(a1+a2)]x[1/ λ1]) + ([a2/(a1+a2)]x[1/ λ2]). 0

The parameters a1, a2, λ1, and λ2 were obtained from double exponential curve-

62 fits of the pulse-escape data using Kaleidagraph software (Synergy Software,

Reading, PA).

2.4.5. Statistical Analysis

Statistical analyses of the velocities and frequencies of neurofilament movement in gaps were performed with SPSS software using the Kolmogorov-Smirnov and

Mann–Whitney tests for two independent samples (SPSS).

Statistical analyses of the pulse-escape data were performed by Drs. X. Li and

Dr. L. Wei from the Center for Biostatistics at The Ohio State University

(Columbus, OH). The raw data were subjected to log transformation to make them more normally distributed, and then the slopes were compared using a linear mixed effects model for repeated measures (Verbecke and Molenberghs,

2000; Diggle et al., 2002). The mixed effects model takes into consideration the fact that the selected axons are sampled from a larger population with inherent differences, like the differences in the initial fluorescence intensities, or experimental differences between neuronal cultures.

The linear mixed effects model also allows analysis for repeated measures, which is most appropriate for a time-series experimental procedure and it considers any missing measurements throughout the time-series as missing at random without affecting the final outcome of the test. We used this model to test the hypothesis that the decay kinetics over time were group-independent

63 (considering the groups to be wild-type and dilute lethal), assuming a first-order autoregressive covariance structure within single pulse-escape traces.

2.5. Immunostaining

Coverslips were washed twice with PBS and fixed for 30 minutes with 4% paraformaldehyde in PBS containing 1% sucrose, after which the cells were treated with 1% Triton X-100 in PBS for 15 minutes and processed for immunostaining using standard procedures. Blocking was performed using 4% normal goat serum (Jackson Immunoresearch).

Neurofilaments were detected using rabbit polyclonal antibody AB1987 (Millipore

Bioscience Research Reagents, Billerica, MA), which is specific for NFM.

Actin was detected using rhodamine phalloidin (Molecular Probes, Invitrogen,

Carlsbad, CA).

As a secondary antibody, we used Alexa 488-conjugated goat anti-rabbit IgG

(Molecular Probes, Invitrogen). Coverslips were mounted on microscope slides using ProLong Gold Antifade reagent (Molecular Probes, Invitrogen).

64 2.6. Western Blotting

Brains and spinal cord tissue from P4 mice were homogenized in SDS-PAGE loading buffer containing 2% SDS using a Teflon glass homogenizer. The homogenates were then sonicated for 5 minutes in a bath sonicator and stored at

-20°C. Protein concentration was determined with the Bradford Protein Assay kit

(Bio-Rad, Hercules, CA) using BSA as a standard.

For electrophoresis, the samples were diluted in loading buffer, heated for 5 minutes at 95°C, centrifuged at 15,000�Xg for 10 minutes at room temperature, and then resolved by SDS-PAGE on 7.5% polyacrylamide gels. Proteins were transferred to PVDF membranes (Millipore Immobilon-P Transfer Membrane, pore size 0.45 µm) by tank blotting.

Myosin Va was detected using rabbit polyclonal antibody LOOP2 obtained from

Dr. John Hammer (Laboratory of Cell Biology, National Institutes of Health/

National Heart, Lung, and Blood Institute, Bethesda, MD).

NFM was detected using a rabbit polyclonal antibody AB1987 (Millipore

Bioscience Research Reagents) or mouse monoclonal antibody RMO270

(Invitrogen) (Lee et al., 1987), both of which bind in a phospho-independent manner (Yan et al., 2007). Phosphorylated NFM was detected using the mouse monoclonal antibody RMO55 received from Dr. Virginia Lee (University of

65 Pennsylvania, Philadelphia, PA) (Lee et al., 1987).

Neurofilament protein H (NFH) was detected using the rabbit polyclonal antibody

AB1989 (Millipore Bioscience Research Reagents), which binds in a phospho- independent manner (Yan et al., 2007). Phosphorylated NFH was detected using mouse monoclonal antibodies SMI34 (Covance, Princeton, NJ) or RT97

(BioDesign, Meridian Life Science, Saco, ME).

α-Tubulin was detected with the mouse monoclonal antibody B-5-1-2 (Sigma).

The secondary Abs were either goat anti-rabbit IgG or goat anti-mouse IgG, each conjugated to HRP (Jackson ImmunoResearch). Blots were processed using

ECL Plus Western blotting detection reagents (GE Healthcare, Piscataway, NJ) and Blue x-ray films (Phenix Research Products, Candler, NC). The films were digitized on a Microtek Scan-maker i900 flatbed scanner (Microtek, Cerritos, CA).

For quantification of neurofilament phosphorylation, the intensities of the corresponding bands from wild-type and dilute lethal samples were measured using MetaMorph software and corrected for background by background subtraction. To avoid concerns about saturation and nonlinearity on the films, we measured the intensity of each band at a range of exposures (5–120 s) and used only those exposures that were in the linear range.

66

Chapter 3: The Role of Myosin Va in Neurofilament Transport

3.1. Introduction

Despite a large body of evidence suggesting that long-range transport of neurofilament polymers takes place along microtubule tracks, carried out by microtubule-dependent molecular motors (Yabe et al., 1999; Shah et al., 2000;

Helfand et al., 2003; Xia et al., 2003; Wagner et al., 2004; Francis et al., 2005;

He et al., 2005), a group of researchers from Ralph Nixon’s lab in 2002 investigated the possible involvement of non-muscle myosin, myosin Va, in axonal transport of neurofilaments, using DLS/LeJ dilute lethal mice. They observed a two-fold increase in the numbers of neurofilament protein in peripheral axons of dilute lethal mice, which lack myosin Va. They also discovered that the actin-dependent motor colocalized and associated with the neurofilaments, as revealed through immunoelectron microscopy and co- immunoprecipitation experiments. Myosin Va levels mirrored those of NFL protein in neuronal axons, as evident by the decrease in myosin Va levels in NFL knock-out mice, and its increase in mice overexpressing NFL (Refer to section

1.4.3 for more details).

67 These observations were all suggestive of a role of myosin Va in neurofilament axonal transport. The increase in neurofilament numbers observed in the absence of myosin Va provided us with three possible models through which the actin-based motor and the neurofilaments are interacting:

1) Neurofilament polymers are accumulating along the axons due to an

increase in the influx rates of neurofilaments transported into the axons

from the cell bodies. This could result from an upregulation of

neurofilament synthesis in the cell, but this explanation lacks a clear

mechanism that links myosin Va to neurofilament production. The second

possibility is the activation of a release signal that allows previously

“anchored” or pausing neurofilaments in the perikarya to be transported in

higher numbers into the axons in the absence of myosin Va. According to

this model, myosin Va would be acting as an anchor to neurofilaments in

the cell body.

2) Neurofilaments, in the absence of myosin Va, are pausing along the axons

for periods of time that are longer than they normally do. As more

neurofilaments get stuck in the pausing state, their numbers increase. In

this model, myosin Va could be regarded as a “facilitator” of neurofilament

transport.

68 3) Neurofilament degradation rates decrease in the absence of myosin Va

while their influx remains constant or does not change as dramatically. As

a result, neurofilament numbers increase in the absence of myosin Va.

There is no previous evidence to suggest that myosin Va could be involved in

the degradation or synthesis of cytoskeletal proteins, while it is very likely that

myosin Va, as a molecular motor, could function in the tethering or the

transport of cargo to their tracks in the cell. In addition, axonal neurofilaments

are long-lived (persisting for several months) under normal conditions in vivo,

and a change in their degradation rates is not likely to alter their axonal

densities in a relatively short period of time (dl mice in the Nixon group study

did not survive beyond 4 weeks of age). Therefore, we hypothesize that

myosin Va plays a role in the axonal transport of neurofilaments, by acting as

a facilitator of neurofilament movement, or an anchor that tethers the

filaments to the cytoskeleton. To test this hypothesis, we investigated the

axonal transport of neurofilaments in cultured neurons from SCG and DRG of

wild type and dilute lethal mice using live-cell imaging.

3.2. Identification of dilute lethal mice from DLS/LeJ and dl20J strains

The DLS/LeJ strain originated from C57BL/Gr strain, which arose from

C57BL/Go strain during the 1950s. C57BL/Go diverged from C57BL strain at the same time as C57BL/6 in 1932. The dl mutation in DLS/LeJ mice was never

69 sequenced, and so it was not possible for us to identify mice from the strain using

PCR genotyping.

It was possible to identify the mice from the litter using phenotypic differences

(Table 2.1). Mice homozygous for the dl mutation showed a lightening in coat color (originally referred to as “Maltese dilution” (Russell, 1949). The coat color phenotype was readily identifiable at four days after birth (P4) (Figure 3.1A).

Wild type and heterozygous mice were dark in color, and showed no difference from each other. At 10 days of age though, ears from wild type mice were shorter than their heterozygous counterparts due to them being homozygous for the Bmp5se mutation maintained in repulsion with dl (refer to section 2.1.2). We confirmed the absence of the myosin Va protein in dl mice using Western blot analysis (Figure 3.1B). Brain tissue from wild type and dilute P4 mice was homogenized and probed for myosin Va. We also probed for NFM and used it as control.

In dl20J mice, the dilute lethal mutation arose spontaneously from C57BL/6 X

DBA2J mice. The mice were back-crossed to C57BL/6 mice and maintained as such. Sequence analysis of the mutation revealed that the dl20J allele is caused by a deletion of 3,580 bp (Strobel et al., 1990) that removes a single coding exon of the dilute gene that renders it as a functional null mutation. It is possible to genotype the mice by PCR run on 2.5% agarose gel in TBE. Wild type mice showed a 334 bp fragment on the gel. Homozygous dl mice showed two bands

70 corresponding to 251 and 291 bp. Heterozygous mice revealed a triplet corresponding to 251, 291 and 3334 bp (Figure 3.1C).

3.3. Neurofilament distribution along neuronal axons of dilute lethal mice

To investigate the role of myosin Va in neurofilament transport, we first utilized live-cell imaging of fluorescently-tagged neurofilaments. We measured the kinetics of transport of these proteins in SCG and DRG neuronal cultures from

DLS/LeJ dilute lethal mice and compared them to neurons extracted from

C57BL/6J wild type mice.

Neurons from mice lacking myosin Va did not show any difference in their development, morphology or survival rates when compared to wild type neurons

(Figure 3.2A, B). Axons from dilute lethal neurons were indistinguishable in length from their wild type counterparts, which supports Evans et al. (1997) observation that myosin Va plays no role in axonal growth. Moreover, the neurofilament distribution did not vary from wild type to dilute lethal mouse neurons, and there was no apparent difference in the number or the distribution of naturally occurring gaps in SCG neurons in our cultures (Figure 3.2C, D).

71 3.4. Short-term mobile behavior of neurofilaments in SCG neurons

We transfected SCG neurons from C57BL/6 wild type and DLS/LeJ dilute lethal

P4 pups with GFP-tagged NFM fusion protein by nuclear injection two days after plating. After 5 days in vitro (DIV), we observed the neurons under a fluorescent microscope to locate the naturally occurring gaps in the neurofilament array of

SCG neurons (~20 nm-long; refer to section 2.4.2 for further details) and collected 15-minute long time-series acquisitions with 4-second intervals and 1- second exposure. These filaments were transported in a rapid fashion, interrupted by pauses that in certain cases extended throughout the observation time (Figure 3.3).

72

Figure 3.1. Phenotyping and genotyping dilute lethal pups A: two P4 DLS/LeJ mouse littermates. Mice lacking myosin Va (-/-) have a lighter coat color than their siblings. B: Western blot analysis of brain tissue from wild type and DLS/LeJ dilute lethal P4 littermates. The upper half of the membrane was probed using anti-myosin Va antibody (LOOP2), and the lower part was probed with NFM antibody (AB1987) as a loading control. C: PCR genotyping of dl20J mice. Wild type mice were identified by the appearance of two bands at 251 and 291 bp. Their dilute lethal littermates were identified by the appearance of one band at 334 bp. Heterozygous mice were identified by the appearance of all three bands.

73

Figure 3.2. Dilute lethal neurons extend axons that contain neurofilaments and gaps A and B: Immunostaining of cultured SCG neurons from wild type and DLS/LeJ dilute lethal mice 24 hours after plating. Neurons from wild type and dilute lethal mice were capable of extending axons with neurofilaments throughout their lengths with no observable difference in their morphology. Scale bar, 20 µm. C and D: There was no apparent difference in the number or distribution of gaps in the neurofilament array between wild type and dilute lethal axons. Scale bar, 4 µm.

74

Figure 3.3. Examples of moving neurofilaments in wild type and dilute lethal axons A, B, C: trajectories of three neurofilaments in wild type axons. D, E, F: trajectories of three neurofilaments in dilute lethal axons. Each point in the graphs represents the location of the neurofilament in one 4-second time interval. The x-axis represents time elapsed from the beginning of the movie, and the y-axis corresponds to the distance, in µm, from the location the neurofilament tracking started. Neurofilaments in A and D moved anterogradely, in B and E retrogradely, and in C and F, they moved in one direction, then reversed to move in the opposite direction.

75 3.4.1. Neurofilament length and frequency of movement in naturally occurring gaps We were able to track the movement of 58 neurofilaments in wild type SCG neurons and 53 neurofilaments in DLS/LeJ neurons. Each neurofilament had a preferred direction of movement along the axon, allowing us to classify them as anterograde or retrograde based on their net directionality of transport (Figure

3.3). We also observed few neurofilaments that reversed their direction of transport along the axons from both wild type neurons (three reversals) and dilute lethal neurons (two reversals).

We calculated the frequency of GFP-tagged neurofilament polymers moving through the gaps. The average length of tracked neurofilaments was 3.9 µm in the wild type and 4.6 µm in the dilute lethal. Statistical analysis using the

Kolmogorov-Smirnoff (KS) and the Mann-Whitney tests showed no statistically significant difference between the two populations. We have no reason to expect a difference in neurofilament length, either.

The average frequency in wild type was 0.22 filaments/minute (an average of 3.3 filaments per 15-minute movie). In dilute lethal (DLS/LeJ) neurons, the average frequency decreased by 27% to 0.16 filaments/minute (an average of 2.4 filaments per 15-minute movie). Using the KS and Mann-Whitney tests, we found that this difference did not rise to statistical significance (Figure 3.4; Table

3.1).

76

Filament Filament Frequency of Average velocity Average velocity Average peak number length movement including pauses excluding pauses velocity (µm) (fil/min) (µm/sec) (µm/sec) (µm/sec)

wild type 24 3.7 0.09 0.17 0.36 0.79

dilute lethal 25 4.4 0.07 0.29 0.47 0.89

P-value (K-S test) 0.223 0.290 0.704 0.938 0.867

Significant No No No No No Anterograde difference

wild type 34 4.0 0.13 -0.32 -0.49 -1.13

dilute lethal 28 4.8 0.09 -0.22 -0.41 -1.23

P-value (K-S test) 0.575 0.356 0.876 0.659 0.245

Significant No No No No No difference

Retrograde

Table 3.1. Statistical comparison of neurofilament movement in wild type and dilute lethal neurons Average length, frequency and velocity measurements for 58 moving neurofilaments in wild type neurons and 53 moving neurofilaments in dilute lethal neurons. For neurofilaments that exhibited a reversal of direction, the anterograde and retrograde excursions were treated as separate anterograde and retrograde movements. For statistical analysis we used the Mann-Whitney and Kolmogorov-Smirnov (K-S) tests and only values from the K-S test are shown here.

3.4.2. Kinetics of neurofilament transport along mouse SCG neurons

To test our hypothesis that the absence of myosin Va results in a change in the

transport kinetics of neurofilament polymers, we measured the velocities of

neurofilament movement in wild type and dilute lethal neurons. The average

velocity of neurofilament transport in the anterograde direction increased in the

absence of myosin Va from 0.17 µm/second to 0.29 µm/, but this difference was

not statistically significant. The average velocity in the retrograde direction

decreased from 0.32 µm/second to 0.22 µm/second in the absence of myosin

77 Va, but this difference was not statistically significant either (Figure 3.4; Table

3.1).

We also looked at the average velocity excluding pauses to allow us a closer look at the motor velocities, unbiased by the durations of the pausing intervals.

Similar to the previous velocity measurements, we observed an increase in the average velocities in the anterograde direction (from 0.36 to 0.47 µm/second) and a decrease in the average velocities in the retrograde direction (from 0.49 to

0.41 µm/second). We failed to observe a statistically significant difference in these measurements using either the KS or the Mann-Whitney tests (Figure 3.4;

Table 3.1).

We also compared the average peak velocity of transport from wild type and dilute lethal mice. These measurements showed an increase in transport velocities in both the anterograde (from 0.79 µm/second in wild type to 0.89

µm/second in dilute lethal) and the retrograde directions (from 1.13 µm/second in wild type to 1.23 µm/second in dilute lethal). Again, we failed to observe a statistically significant difference between the two populations using the KS and

Mann-Whitney tests (Figure 3.4; Table 3.1).

These results indicate that, although there is a difference between the frequencies and kinetics of neurofilament transport from wild type and dilute lethal mice, the difference is not dramatic enough to rise to statistical

78 significance. This does not necessarily imply that myosin Va plays no role in neurofilament transport, because the possibility remains that myosin Va could affect the long-term pausing behavior of neurofilaments stranded outside the naturally-occurring gaps, which is not possible to study using 15-minute movies of mobile neurofilaments. Instead, we need to utilize an approach that explores the long-term pausing of neurofilaments, and how the duration of pauses varies in the presence or absence of the myosin Va from the neurons.

79

Figure 3.4. Analysis of neurofilament movement in wild type and dilute lethal axons Histograms of frequency of movement (A, B), average velocities (C-F), and peak velocities of neurofilaments in wild type (n=58) and DLS/LeJ dilute lethal (n=53) axons. Anterograde and retrograde movements are represented on the y-axis as positive and negative values, respectively. Statistical analysis of the data showed no significant difference between wild type and dilute lethal for either direction of movement (Table 3.1).

80

81 3.5. Long-term mobile behavior of neurofilaments in DRG neurons

In Trivedi et al. (2007), the authors developed a novel “pulse-escape fluorescence photoactivation” technique, which allowed them to observe a large population of stationary neurofilaments that paused for prolonged periods of time along axons of rat SCG neurons. They estimated that in these neurons, neurofilaments spend ~97% of their time pausing along the axons, while only

~3% of the neurofilament polymers are moving at any given point in time. They estimate that the pausing duration of stationary neurofilaments is 60 minutes, much longer than the duration of our 15-minute movies. Such a result clearly highlights the limitation of our 15-minute time-lapse movies, which allow us to analyze the transport kinetics of a small population of neurofilaments that move through the short gaps, but it offers no insight on >97% of the neurofilaments that pause outside these gaps.

For this reason, we utilized the pulse-escape fluorescence photoactivation technique to compare the long-term pausing behavior of axonal neurofilaments from wild type and dilute lethal mice. As mentioned earlier (section 2.4.2), the low neurofilament content along SCG axons gives rise to the appearance of gaps. The high neurofilament content in mouse DRG neurons, on the other hand, gives rise to a continuous neurofilament array along the axons. This makes them a better model to observe the long-term pausing of stationary neurofilaments.

82 3.5.1. Long-term pausing in DRG neurons from DLS/LeJ mice

DRG neurons from P4 dilute lethal mice of the DLS/LeJ strain that was used by

Rao et al. (2002) were transfected after 3 DIV with PAGFP-NFM and DsRed2 and observed at 5 DIV. As controls, we used wild type C57BL/6 mice of the same age, and treated similarly as their dilute lethal counterparts.

The green fluorescence along axons from transfected cells was activated in ~20

µm segments using a beam of violet light with 1000-msecond exposures.

Fluorescence intensities in these regions at time= 0 seconds post-exposure was used as the starting fluorescence intensity. Time-lapse imaging with 5-minute intervals for a period of 115 minutes allowed us to track the rate of departure of neurofilaments from the activated regions in axons from wild type and dilute lethal mice (Figure 3.5).

In both groups, the decrease in fluorescence was gradual and many neurofilaments remained in the activated regions after 115 minutes, as expected due to the long-term pausing of neurofilaments along the axons. For wild type

DRG neurons, the average intensity remaining in the activated region after 115 minutes was 65% of the starting intensity. The average intensity remaining in the activated region of neurons from DLS/LeJ DRG was 73% of the starting intensity

(Figure 3.6).

83 Similar to the results from Trivedi et al. (2007), our data were biphasic and fit a double-exponential function, indicating that there are two distinct populations of neurofilaments, one that pauses for short periods of time (the first filaments to leave the activated region; Figure 3.6), and a second population with much longer pauses. Calculations of the mean lifetime, or mean time for departure of neurofilaments from the activated region, offers a better appreciation of the difference in the kinetics between filaments from wild type and dilute lethal mice

(for detailed equations, refer to section 2.4.4). Using the equation generated by the double exponential curve fits, we found that the mean time for departure of neurofilaments from wild type mice was 315 minutes. The time was prolonged significantly up to 466 minutes in neurons from dilute lethal DRG cultures. This

48% increase in the time to depart in dilute lethal showcases the large difference in transport kinetics that becomes apparent when considering longer time scales, which was not possible to observe using the previous 15-minute observations in

SCG neuronal cultures.

84

Figure 3.5. A pulse-escape fluorescence photoactivation experiment An example of a wild type DRG neuron that was transfected with PAGFP-NFM and a diffusible red fluorescent potein (DsRed2) as a marker for transfection. A: The DsRed2 fluorescence marks the transfected axon. B: The same region prior to activation show no green fluorescence. C: The same region immediately after activation using violet light. D: 1 hour after activation, the same regions has lost some of its fluorescence. E: 115 minutes after activation, fluorescence intensity in the activated region decreases further, the decrease corresponds to fluorescent neurofilaments leaving the activated region, although many of them remain. Arrow heads mark the location of the activated region. Scale bar: 5µm.

85 3.5.2. Statistical analysis of pulse-escape data from DLS/LeJ mouse neurons

To analyze the data statistically, it would be misleading to compare the fluorescent intensities that remain in the activated regions after 115 minutes, because the kinetics of transport are affected by every frame within the time- series. To compare the different profiles we performed logarithmic transformations on the raw data of each pulse-escape profile from wild type and dilute lethal neurons. This transformation resulted in linearizing the data, making it possible to calculate the slope for every time-series and compare the average slopes from wild type to those from dilute lethal mice using a linear-mixed-effects model for repeated measures, assuming a first-order autoregressive covariance structure within each data set and random effects for the slopes and intercepts to account for variability between axons (for details refer to section 2.5). The average slope was -0.019 ln[ADU*µm-1]*min-1 for the wild-type axons and -0.013 ln[ADU*µm-1]*min-1 for the dilute lethal axons, which represents a 32% decrease

(Figure 3.6), and this was statistically significant (p<0.001).

3.5.3. Long-term pausing in DRG neurons from dl20J mice

As mentioned earlier, the genetic background of the dilute lethal mice of DLS/LeJ is different from that of C57BL/6 wild type mice used as controls. One of the concerns that we had as a result was the possibility that this difference in background could affect the observed difference in neurofilament pausing between the two populations.

86 To address this concern, we decided to repeat the pulse-escape experiment using mice from dl20J strain, where the dilute lethal mutation has been sequenced. This strain is the best-characterized dilute lethal strain with a functional null dilute gene. The advantage of using dl20J mice is our ability to genotype the litters and, therefore, identify homozygous dilute lethal and wild type littermates that can be used as control that share the same genetic background.

The average intensity remaining in the activated region from wild type after 115 minutes was 69% of the starting intensity. The mean time for departure from these regions was 482 minutes. The average intensity remaining in the activated region of neurons from dilute lethal was 88% of the starting intensity, and the mean time for departure was 1294 minutes. This accounts for a 169% increase compared with wild type (Figure 3.7).

Statistical analysis of the data using the linear-mixed effects model described above yielded average slopes of -0.013 ln[ADU*µm-1]*min-1 for wild type and

-0.005 ln[ADU*µm-1]*min-1 for dilute lethal axons (Figure 3.7). This difference was highly statistically significant (p<0.0001).

87

Figure 3.6. Pulse-escape kinetics for wild type and DLS/LeJ mice A, B: box and whisker plots for 27 wild type axons and 24 DLS/LeJ dilute lethal axons that were imaged for 115 minutes at 5-minute intervals. The axon-to-axon variability is due in part to the stochastic nature of the movement. C,D: plots of the average fluorescence intensities. After 115 minutes the average percentage of the initial fluorescence remaining in the activated regions was 65% for the wild type and 73% for the dilute lethal. The data matches a double-exponential decay (curve-fit). Considering the pulse-escape kinetics to represent a probability density function, the mean time for a neurofilament to depart the activated regions was equal to 5.3 hours in wild type neurons and 7.8 hours in dilute lethal neurons. E-J: plots the log-transformed kinetics for three representative wild type axons and three dilute lethal axons. K- L: histograms of th slopes for 27 wild type and 24 dilute lethal neurons.

88

Figure 3.7. Pulse-escape kinetics for wild type and dl20J mice A, B: box and whisker plots for 23 wild type axons and 24 dl20J dilute lethal axons that were imaged for 115 minutes at 5-minute intervals. The axon-to-axon variability is due in part to the stochastic nature of the movement. C,D: plots of the average fluorescence intensities. After 115 minutes the average percentage of the initial fluorescence remaining in the activated regions was 69% for the wild type and 88% for the dilute lethal. The data matches a double-exponential decay (curve-fit). Considering the pulse-escape kinetics to represent a probability density function, the mean time for a neurofilament to depart the activated regions was equal to 8.0 hours in wild type neurons and 21.6 hours in dilute lethal neurons. E-J: plots the log-transformed kinetics for three representative wild type axons and three dilute lethal axons. K-L: histograms of the slopes for 23 wild type and 24 dilute lethal neurons.

89 3.6. Investigating neurofilament phosphorylation state in dilute lethal mice

As mentioned earlier, it is proposed that the phosphorylation of neurofilament subunits, especially KSP sites of NFM and NFH tails could result in an overall decrease in neurofilament transport caused by an increase in the pausing times

(for details refer to section 1.2.5). Therefore, one possible explanation for the prolonged pauses observed in the absence of myosin Va in dilute lethal mice, is an increase in neurofilament phosphorylation.

To test this hypothesis, we explored the phosphorylation state of NFM and NFH using quantitative Western blot analysis with phospho-dependent and phospho- independent antibodies. Spinal cords from P4 wild type and dilute lethal mice were extracted, homogenized and probed using RMO270, an antibody that binds

NFM independently of its phosphorylation state (Lee et al., 1988b). RMO55 was used to probe for phosphorylated NFM (Lee et al., 1987)Brown, 1998). To probe for NFH, we used pAB1989, which can recognize both phosphorylated and non- phosphorylated NFH (Harris et al., 1993). SMI34 is a phosphorylation-dependent antibody (Lichtenberg-Kraag et al., 1992) that we used to probe for phospho-

NFH in addition to RT97. This latter has been well-characterized in a recent study (Veeranna et al., 2008) that reported the generation of RT97 epitopes by phosphorylation at KSPXK and KSPXXXK sites on NFH tails.

Our quantification of the blot staining intensities (methods in section 2.7) shows no significant difference in the amount of total or phosphorylated NFM and NFH

90 protein between wild type and dilute lethal samples (Figure 3.8). This indicates that the increase in pausing behavior observed earlier is not caused by neurofilament hyperphosphorylation.

Figure 3.8. Neurofilament phosphorylation state in wild type and dl20J mice A: Western blots of mouse spinal cords homogenates from wild type (wt) and dl20J dilute lethal (dl) P4 mice. The upper half of the blot was probed with neurofilament antibodies, whereas the lower half was probed with tubulin antibody (α-tub). NFM was detected with monoclonal antibody RMO270, which binds in a phospho-independent manner. Phosphorylated NFM was detected with monoclonal antibodies RMO55, which binds to phosphorylated epitopes on NFM in a phospho-dependent manner. NFH was detected with polyclonal AB1989, which binds in a phospho-independent manner. Phosphorylated NFH was detected with mouse monoclonal antibodies SMI34 and RT97, which bind to phosphorylated epitopes on NFH in a phospho- dependent manner. Tubulin served as a loading control and was detected with monoclonal B-5-1- 2. B: Quantification of blot staining intensities. Each blot was performed in triplicate. For each lane on each blot, the background-corrected intensity of the neurofilament band was divided by the background-correlated intensity of the corresponding tubulin band. For each neurofilament antibody, the resulting intensity ratios were then averaged and normalized to the average wild type. The error bars represent the SD about the mean. We observe no significant difference in the intensities of the bands in wild type and dilute lethal tissues. (p=0.95 for RMO270; p=0.67 for RMO55; p=0.73 for AB1989; p=0.96 for RT97; p=0.97 for SMI34; Student’s t-test)

91 3.7. Summary

In summary, the present data reveals that in the absence of myosin Va, neurofilaments spend considerably more time pausing along neuronal axons compared to controls. In SCG neurons from dilute lethal mice, we report a 27% decrease in the frequency of neurofilament axonal transport. Yet, the change in frequency and velocity of transport did not rise to statistical significance. This subtle effect indicates that myosin Va does not play a major role in long-range axonal transport of neurofilaments but the possibility remains that myosin Va could affect the long-term pausing behavior of neurofilaments that are not observed using 15-minute movies of filaments moving through naturally-occuring gaps.

To study the role of myosin Va in long-term pausing of neurofilaments, we used the pulse-escape fluorescent photoactivation technique, which revealed that in two different strains of dilute lethal mice, neurofilaments pause for significantly longer periods of time compared to wild type controls. The increase in pausing durations was found to be independent of phosphorylation state of NFM or NFH subunits that are proposed to regulate neurofilament transport. Therefore, we propose that myosin Va facilitates the transport of neurofilaments along neuronal axons by decreasing the pausing durations of neurofilaments. In the absence of myosin Va, neurofilaments accumulate along the axons, similar to the observation in DLS/LeJ dilute lethal mice reported by the Nixon group (Rao et al.,

92 2002). As neurofilaments pause for prolonged periods of time, their numbers increase resulting in their accumulation along the axons, as observed in vivo.

93

Chapter 4: The Role of the Dynein/Dynactin Complex in Neurofilament Transport

4.1. Introduction

It has become the general assumption that the MT-based motor proteins are responsible for neurofilament long-range axonal transport. This view was strengthened after live-cell imaging of fluorescently tagged neurofilaments displayed rapid and intermittent bouts of movement at speeds close to those of dynein and kinesins (Roy et al., 2000; Wang et al., 2000). Major contributions to the debate came from two studies that were able to show direct interaction between neurofilaments and the dynein/dynactin complex in vivo and in vitro.

The first study, by Shah et al. (2000), neurofilament polymers isolated from bovine and rat spinal cord and brain were associated with dynein/dynactin. The bidirectional transport of the isolated filaments was reconstituted in vitro along microtubule tracks. The dynein/dynactin complex was disrupted using antibodies directed against the dynein intermediate chain, and dissociated neurofilaments from dynein, but not from dynactin. These results, along with immuno-electron microscopy observations, reveal a direct interaction between neurofilaments and

94 the dynein/dynactin complex, and that dynein/dynactin bound to the isolated neurofilaments are capable of moving the polymers on microtubule tracks in vitro.

Yet, this result does not prove that dynein/dynactin is the motor responsible for retrograde neurofilament movement in vivo.

In 2002, a second study from Erika Holzbaur’s lab (Wagner et al., 2002) showed, using atomic force microscopy (AFM) and biochemistry that the dynein/dynactin complex isolated from rat brain tissue could bind to neurofilament polymers. The interaction was inhibited using dynein intermediate chain antibody. The authors were also able to isolate a subclone of NFM bound to dynein intermediate chain and confirmed the specificity of the interaction using affinity-chromatography.

This result shows that dynein/dynactin can bind neurofilaments in vivo through the association of NFM with dynein intermediate chain, but, like Shah et al.

(2000), neurofilament transport was not measured directly, and the study fails to confirm that dynein is the retrograde motor of neurofilaments.

Earlier studies in our lab have demonstrated the dependence of neurofilament transport in cultured neurons on kinesin-1A, B and C (Uchida et al., 2009). Using kinesin-1A knock out mice, Uchida et al. calculated the anterograde and retrograde frequencies of neurofilament transport along SCG neuronal axons.

They report a decrease in anterograde movement that was surprisingly mirrored by a decrease in retrograde movement. Transfection of the cells with a kinesin-

1A construct rescued transport in both directions, and transfection with kinesin-

95 1B and C resulted in partial rescue in both directions as well, suggesting a functional redundancy between kinesin-1 isoforms. To confirm the result, SCG neurons from wild type mice were transfected with headless kinesin-1A isoforms.

This also caused a disruption of neurofilament transport in both the anterograde and retrograde directions. The data from this study suggests a mechanistic inter- dependence between the two oppositely directed motors and a possible coupling their activities such that the activity of kinesin-1A is required for the activity of dynein. This raises the question of the dependence of kinesin motor on the activity of the dynein/dynactin complex.

The only study to look into the role of dynein/dynactin in axonal transport of neurofilaments came from Peter Baas’s group (He et al., 2005). The authors disrupted dynein using dynein heavy chain siRNA in rat SCG neurons. This caused a dramatic decrease in the frequency of retrograde movements, whereas the frequency of anterograde movements increased. The authors did not look into the kinetics of transport, and did not confirm the result using a different approach to disrupt dynein activity, probably since the primary focus of the study was not neurofilament transport, but rather microtubule axonal transport. Yet, this result clearly does not suggest a reciprocal interdependence between dynein and kinesin activity.

96

Figure 4.1. Efficiency of DHC siRNA knock down after 4 and 7 days in culture Efficiency of knock down determined by immunostaining using dynein heavy chain antibody. Control: cells not injected with siRNA (n=10 immunostained cells). 4DIV: cells injected with DHC siRNA and immunostained after 4 days in culture, 3 days after siRNA injection (n=10) showed a decrease in DHC fluorescence by 24%. 7DIV: cells injected with DHC siRNA and immunostained after 7 days in culture, 6 days after siRNA injection (n=10) revealed a 76% decrease in DHC fluorescence.

97

Figure 4.2. Effect of dynein heavy chain knock down on neurofilament movement Control: Untransfected SCG neurons (5-7 DIV; 59 movies). Control siRNA: cells injected with scrambled with scrambled siRNA observed 6 days after injection (7DIV; 23 movies). Frequency of neurofilament transport was not affected in either direction. DHC siRNA: cells injected with a pool of three siRNAs targeting dynein heavy chain (20 movies). Neurofilament transport in both the retrograde and anterograde directions were significantly decreased. The data was statistically compared to control using Mann-Whitney test. P-values: ***: p<0.001; **: p<0.01; *: p<0.05.

4.2. Knock-down of dynein heavy chain in SCG neuronal cultures

To confirm the result obtained from He et al. (2005), we disrupted dynein function in mouse SCG neurons using siRNA oligonucleotide duplexes targeting non- overlapping regions of DHC (Stealth™ Select 3 RNAi; Invitrogen). To measure the efficiency of knock-down, we fixed the cells, immunostained the heavy chain and measured the fluorescence intensity in cells transfected with siRNA and compared them to non-transfected cells (Materials section 2.6). We found that

98 reduction in protein levels starts at Day 4 after plating, with 24% decrease in

DHC expression in injected cells (n=10) as compared to control (Figure 4.1). The efficiency of knock-down increases on day 7 after plating. We report 76% reduction in dynein heavy chain expression levels in transfected cells when compared to non-transfected controls as deduced from immunostaining (n=10;

Figure 4.1). We used the corresponding scrambled Stealth™ RNAi Negative

Control Duplex with medium GC content (Invitrogen) to confirm the specificity of knock-down. In these cells, dynein heavy chain levels were 102% compared to non-transfected cells (n=10), which confirms the specificity of our dynein heavy chain siRNA targeting (Figure 4.1).

Neurons transfected with dynein heavy chain siRNA 24 hours after plating were able to extend axons with a normal neurofilament array. We were able to observe the movement of tagged polymers as they moved through naturally- occurring gaps from 20 axons in 7DIV-neurons. In cells transfected with the scrambled control siRNA template 24 hours after plating (n=23), there was no significant difference in the transport frequencies (7DIV cells) as compared to non-transfected cells (Tables 4.1, 4.2 and Figure 4.2 (n=59; 99% and 103% in anterograde and retrograde transport, respectively). Based on these observations, we calculated 54% reduction in the retrograde transport of neurofilaments as compared to controls injected with a scrambled siRNA construct from 23 axons (Table 4.1; Figure 4.2). Surprisingly, we also observed a 66% decrease in the transport frequency of anterogradely moving filaments

99 from movies (Table 4.1; Figure 4.2). These results contradict the observations

from He et al. (2005) and suggest a role for dynein/dynactin in anterograde

transport of neurofilaments. For this reason, we tried to confirm our observations

using a number of approaches previously known to disrupt dynein/dynactin

activity.

Control siRNA DHC siRNA Control Ab DIC Ab Control

Antero Retro Antero Retro Antero Retro Antero Retro Antero Retro

filaments/min 0.069 0.087 0.023 0.040 0.065 0.063 0.010 0.017 0.072 0.085 Frequency filaments/movie 1.04 1.30 0.35 0.60 0.97 0.95 0.15 0.25 1.08 1.27

Number of filaments 24 30 7 12 37 36 3 5 64 75

Total number of filaments 47 18 69 8 125

Number of movies 23 20 38 20 59

Total observation time (min) 345 300 570 300 884

Table 4.1. Neurofilament transport along SCG neuronal axons in culture Control siRNA: cells injected with a scrambled siRNA construct 24 hours after plating. Control Ab: cells injected with mouse non-specific antibody. Control: non-injected SCG neuronal cells.

100

Control Control siRNA DHC siRNA Control Ab DIC Ab Antero Retro Antero Retro Antero Retro Antero Retro Antero Retro filaments/min 0.072 0.085 0.069 0.087 0.023 0.040 0.065 0.063 0.010 0.017 Frequency filaments/movie 1.08 1.27 1.04 1.30 0.35 0.60 0.97 0.95 0.15 0.25 vs. Control 0.000 0.000 0.917 0.962 0.034 0.042 0.463 0.171 0.001 0.001 vs. Control siRNA 0.917 0.962 0.000 0.000 0.038 0.035 n.d. n.d. n.d. n.d. vs. DHC siRNA 0.034 0.042 0.038 0.035 0.000 0.000 n.d. n.d. n.d. n.d. p-value vs. Control Ab 0.463 0.171 n.d. n.d. n.d. n.d. 0.000 0.000 0.006 0.014 vs. DIC Ab 0.001 0.001 n.d. n.d. n.d. n.d. 0.006 0.014 0.000 0.000 vs p50 0.017 0.001 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Number of 64 75 24 30 7 12 37 36 3 5 filaments Total number of filaments 125 47 18 69 8 Number of 59 23 20 38 20 movies Total observation 884 345 300 570 300 time (min)

Table 4.2. Summary of statistical analysis for neurofilament movement in dynein inhibition experiments using DHC siRNA and DIC 74.1 Ab Summary of the p-values obtained in our statistical comparisons of the frequency of neurofilament movement using Mann-Whitney test.

4.3. Functional inhibition of DIC in SCG neurons

In their study, Shah et al. (2000), were capable of inhibiting neurofilament

transport along microtubule tracks in vitro using a function-blocking dynein

intermediate chain antibody (74.1) as mentioned earlier, but they did not directly

observe neurofilament transport along neuronal axons. Treatment with the

antibody (74.1) dissociated neurofilaments from dynein, but not from p150Glued,

suggesting that neurofilament/dynactin complex remains intact.

Using cytoplasmic injections of the same antibody (74.1), we investigated the

effect of dynein intermediate chain function inhibition on neurofilament transport

in 5DIV-mouse SCG neurons and compared the results to control neurons

101 injected with non-specific mouse IgG. Neurofilament transport frequency in cells injected with the non-specific mouse IgG decreased by 8% and 25% in the anterograde and retrograde directions, respectively (Table 4.1), but the difference between non-injected and injected cells did not rise to statistical significance for either direction (Table 4.2). The frequency of retrograde transport 2-6 hours after (74.1) antibody injection was reduced by 80% compared to controls injected with the non-specific antibody, decreasing from an average of

0.95 (n=38 movies) to 0.25 (n=20 movies) filaments per 15-minute movie.

Interestingly, we also observed a decrease in anterograde transport frequency from 0.97 to 0.15 filaments per 15 minutes, corresponding to 86% reduction

(Figure 4.3; Table 4.1). This result confirms our earlier observation using dynein heavy chain knock-down, suggesting a reciprocal interdependence between the anterograde and retrograde motors. To confirm those results using a dominant negative strategy, we used a third approach to disrupt dynein/dynactin function: overexpression of dynamitin/p50 in the cells dissociates the dynactin complex, which is the dynein activator, to inhibit retrograde transport (Echeverri et al.,

1996; Melkonian et al., 2007).

102

Figure 4.3. Effect of function-blocking dynein intermediate chain antibody (74.1) on neurofilament transport Cells were observed 2-6 hours after antibody injection, five days after plating. Control: Un- injected SCG neurons (5-7 DIV; 59 movies). Control IgG: cells injected with a non-specific mouse IgG observed 2-6 days after injection (5DIV; 38 movies). DIC Ab: cells injected with dynein intermediate chain antibody (20 movies). The data was statistically compared to control using Mann-Whitney test. P-values: ***: p<0.001; **: p<0.01; *: p<0.05.

4.4. Dynamitin/p50 overexpression in SCG neurons

Dynamitin/p50 is part of the heterotrimeric shoulder complex of dynactin (Figure

1.4). Upon its overexpression, dynamitin causes the dissociation of p150Glued from Arp1, without affecting dynein integrity (Melkonian et al., 2007). We transfected SCG neurons with GFP-tagged chicken dynamitin. The viability of the cells was comprised in these cultures and we were not able to observe GFP fluorescence in healthy cells. For this reason, we used a myc-tagged chicken

103 dynamitin construct, obtained from Dr. Richard Vallee. Again, the viability of the cells in these cultures was poor, we were able to find healthy, transfected cells.

Mobley and his colleagues (Wu et al., 2007) propose that neuronal cell death after dynamitin overexpression results from the inhibition of retrograde neurotrophin transport, as deduced from a study on PC12 and DRG neurons in culture. Our measurement of the neurofilament transport frequency in healthy

SCG neurons transfected with dynamitin shows a reduction from 1.27 filaments per 15-minute movie from 59 movies to 0.48 filaments per 15-minute movie from

44 movies in the retrograde direction (63% decrease). This disruption of neurofilament movement in the retrograde direction was mirrored by a disruption in anterograde movement, which decreased from 1.08 filaments per 15-minutes to 0.48 filaments per 15 minutes corresponding to 56% decrease (Figure 4.4A).

The larger sample size of moving neurofilaments along axons overexpressing dynamitin as compared to the number of moving neurofilaments in previous experiments (heavy chain knock-down and intermediate chain antibody treatment) allowed us to calculate the velocities of movement, and compare them to velocities in control neurons using statistical analysis. We found that the net average velocity including pauses from cells overexpressing dynamitin was 0.32

µm/second (n=21 filaments) and 0.36 µm/second (n=21 filaments) in the anterograde and retrograde directions, respectively, as compared to 0.37

µm/second (n=64) and 0.31 µm/second (n=75) in controls (Figure 4.4B).

104 Statistical analysis showed no significant difference between the net average velocity of neurofilament movement between the two populations.

The movement of neurofilaments along the axons is interrupted by short and long-term pauses, which contribute to the net average velocities observed. To exclude the contribution of pause durations on these averages, we also calculated the average velocity excluding pauses. In cells overexpressing dynamitin, the velocities were 0.58 µm/second and 0.51 µm/second in the anterograde and retrograde directions, respectively, compared to 0.52

µm/second and 0.60 µm/second in control (Figure 4.4B). Statistical analysis showed no significant difference between the average velocity of neurofilament movement excluding pauses between the two populations. These results suggest that the pause durations are comparable in both populations of observed filaments.

We also calculated the average peak velocities for moving neurofilaments in cells overexpressing dynamitin and control. The average peak velocity in cells overexpressing dynamitin was 0.97 µm/second and 1.19 µm/second in the anterograde and retrograde directions, respectively. This was not significantly different from neurofilament average peak velocity of neurofilaments from control cells, which were 0.91 µm/second and 1.47 µm/second in the anterograde and retrograde directions, respectively (Figure 4.4B).

105 Control p50

Antero Retro Antero Retro filaments/min 0.072 0.085 0.032 0.032 Frequency filaments/movie 1.08 1.27 0.48 0.48

Number of filaments 64 75 21 21

Total number of filaments 125 40

Number of movies 59 44

Total observation time (min) 884 660

Table 4.3. Neurofilament transport along SCG neuronal axons in culture: control vs p50 Control: non-injected cells (5-7 DIV). p50: cells transfected with p50 construct were observed after 5 days of plating (3 days after injection).

106

A. B.

Figure 4.4. Summary of results after p50 overexpression in SCG neuronal cells A. Cells were observed 5-7 days after plating. Control: untransfected SCG neurons (5-7DIV; 59 movies). P50 overexpression: cells injected with dynamitin/p50 (5DIV; 44 movies). B. Summary of the kinetics of neurofilament transport in untransfected cells (Control) and cells overexpressing dynamitin/p50 (p50). The data was statistically compared to control using Mann-Whitney test. P- values: ***: p<0.001; **: p<0.01; *: p<0.05.

4.5. Overexpression of coiled coil 1 domain of p150 in SCG neurons

P150Glued is a dynactin subunit that acts as a binding partner of both microtubules and dynein (Figure 1.4; Karki and Holzbaur, 1995; Vaughan and Vallee, 1995;

Waterman-Storer et al., 1995; Quintyne et al., 1999; Vaughan et al., 2001).

Overexpression of full-length p150Glued and a p150Glued fragment, p150217-548 (also known as coiled-coil 1; CC1) have no effect on endogenous dynactin structure, but act as competitive inhibitors of dynein by associating with the motor and preventing it from binding dynactin and cargo (Quintyne and Schroer, 1999 and

2002).

107

We introduced the DsRed-p150-CC1 construct into 2 DIV mouse SCG neurons by nuclear injections. At 5 DIV, we studied neurofilament transport through naturally-occurring gaps. As expected, we observed a decrease in retrograde transport to 9% of control (from 1.27 to 0.12 filaments per 15-minute movie).

Interestingly, we observed 65% increase in the frequency of anterograde transport (from 1.08 to 1.82 filaments per 15-minute movie), which does not agree with the effect of dynein disruption on anterograde transport we observed earlier (Figure 4.5 and Table 4.4).

We also examined the kinetics of movement of anterograde filaments (n=62).

The net average velocity was 0.22 µm/second (n=62) as compared to 0.37

µm/second in control (n=64). Average velocity excluding pauses was 0.43

µm/second compared to 0.58 µm/second in control, and the average peak velocity was 0.89 µm/second compared to 0.91 µm/second in control. None of these velocities was significantly different from wild-type controls, suggesting that inhibition of the retrograde motor does not decrease the load on the anterograde motor. This observation does not agree with the predictions of a simple tug-of- war model which is discussed in detail in section 5.2.

Control p150-CC1

108 Antero Retro Antero Retro filaments/min 0.072 0.085 0.120 0.008 Frequency filaments/movie 1.08 1.27 1.82 0.12

Number of filaments 64 75 62 4

Total number of filaments 125 65

Number of movies 59 34

Total observation time (min) 884 510

Table 4.4. Neurofilament transport along SCG neuronal axons in culture: control vs p150- CC1 Control: non-injected cells (5-7 DIV). p150-CC1: cells transfected with the coiled coil domain of p150 (5DIV).

Figure 4.5. Frequency of neurofilament transport after p150-CC1 overexpression in SCG neurons Cells injected with p150-CC1 were observed 5 days after plating. Control: untransfected SCG neurons (5-7DIV; 59 movies). p150-CC1: cells injected with the coiled coil1 domain of p150 (5DIV; 34 movies). The data was statistically compared to control using Mann-Whitney test. P- values: ***: p<0.001; **: p<0.01; *: p<0.05.

109 4.6. Overexpression of coiled coil 1 domain of p150 in cortical neurons

The reason for the different result observed in neurons overexpressing p150-

CC1 from other manipulations could offer further insight into the differential roles played by different subunits in the coordination of molecular motors. To investigate and confirm the effect of p150-CC1 overexpression in a different neuronal cell type, we used cultured cortical neurons extracted from E16.5 ICR mice, transfected them with varying concentrations of the expression vector prior to plating, and monitored neurofilament movements along the axons at different time-points.

In cells transfected with 25 µg/ml of the expression construct, we observed a reduction in anterograde and retrograde frequency of 31% and 46% respectively after two days in culture. After four days in culture, the effect of p150-CC1 overexpression was more dramatic, causing a reduction of 75% and 81% respectively in anterograde and retrograde frequency, respectively. The reduction after 2 days in culture was not statistically significant (Figure 4.6; Table

4.5).

We tried to test the possibility that the disruption of anterograde and retrograde transport frequency was concentration-dependent. For this reason, we transfected the cells with 75µg/ml of the p150-CC1 construct. We observed a reduction in anterograde and retrograde frequency of 72% and 66% respectively

110 after two days in culture and a reduction of 70% and 71% respectively after four days in culture (Figure 4.6; Table 4.5).

Control Day2 p150-CC1 Day2 p150-CC1 Day2 (cortical) low conc. high conc.

Antero Retro Antero Retro Antero Retro filaments/min 0.108 0.105 0.058 0.073 0.037 0.030 Frequency filaments/movie 1.62 1.58 0.87 1.09 0.55 0.45

Number of filaments 42 41 20 25 11 9

Total number of filaments 83 45 20

Number of movies 26 23 20

Total observation time (min) 390 345 300

Control Day4 p150-CC1 Day4 p150-CC1 Day4 (cortical) low conc. high conc.

Antero Retro Antero Retro Antero Retro filaments/min 0.130 0.140 0.025 0.035 0.038 0.041 Frequency filaments/movie 1.95 2.10 0.38 0.52 0.57 0.62

Number of filaments 39 42 8 11 12 13

Total number of filaments 81 19 25

Number of movies 20 21 21

Total observation time (min) 300 315 315

Table 4.5. Neurofilament transport along cortical neurons in culture Neurofilament movement was tracked at 2 and 4 days after transfection with 25 µg/ml or 75 µg/ml p150-CC1 construct (labeled “low” and “high” respectively).

111

A. B.

Figure 4.6. Frequency of neurofilament transport after p150-CC1 overexpression in cortical neurons A. Frequency of neurofilament transport in cortical neurons 2 days after plating. Cells transfected with 25 µg/ml of the coiled-coil construct reveal decreased neurofilament movement in both anterograde and retrograde directions, but the decrease was more dramatic in cells transfected with 75 µg/ml of the construct. (Control: n=26; p150-CC1 low: n=23; p150-CC1 high: n=20. B. Frequency of neurofilament transport in cortical neurons 4 days after plating. Cells were transfected with either 25 µg/ml (low) or 75 µg/ml (high) p150-CC1. Neurofilament movement frequency was reduced in both the anterograde and retrograde directions, and it rose to statistical significance as measured by the Mann-Whitney test. P-values: ***: p<0.001; **: p<0.01; *: p<0.05.

112 4.7. Summary

By trying to investigate the role of the dynein/dynactin complex in the axonal transport of neurofilaments, we utilized a variety of techniques to disrupt the motor complex’s function. We targeted a multitude of subunits that are necessary for the activity of retrograde motor, and monitored the effects on neurofilament movement frequency and kinetics.

In all manipulations, we report a total inhibition or a significant decrease in retrograde transport, as expected from disrupting the retrograde motor complex.

This supports hypothesis that dynein/dynactin is the retrograde neurofilament motor. Moreover, in all but one of these targeted disruptions, we observed a reciprocal inhibition of anterograde neurofilament transport, which is also supported by an earlier study from our lab that reported an inhibition of bidirectional neurofilament axonal transport as a result of kinesin inhibition.

Taken together, these studies suggest a tight functional coupling between the dynein/dynactin and kinesin where the activity of the retrograde motor is necessary for the activity of anterograde motor and vice versa.

Despite our confidence in these results, we still have two sources of concern.

The first arises from the fact that our data from the DHC knock-down experiment does not agree with the reported results from He et al. (2005). The cause of the difference is not clear as yet. The second concern we face could hold the answer or a clue as to what is causing the reported discrepancy. Specifically, the

113 overexpression of p150-CC1 in SCG neurons affected anterograde neurofilament transport in a manner opposite to what we observed in cortical neurons where anterograde transport was diminished to the same extent as retrograde transport.

The kinetics of neurofilament transport in SCG neurons were not affected by dynein/dynactin disruption, in spite of the increase in anterograde frequency.

Possible explanations for the contradicting observations could be the difference in neuronal cell type. This possibility is beyond our control, and could arise from internal, inherent differences in the mechanism of motor regulation. Another possibility could be the difference in the introduced construct, which acts as a dominant negative, and different levels of overexpression could produce different effects on the disrupted motor complex. For this reason, we transfected cortical neurons with two different concentrations of the expression construct and compared the effect on transport, but we failed to observe a difference in the effect of motor disruption.

To investigate the possible change in the balance of anterograde to retrograde transport with time, we also monitored transport frequency at different times after transfection. Yet, the disruption of the retrograde motor caused a reduction in anterograde and retrograde transport at two days and four days after transfection, so we do not think that the difference is time-dependent either.

The current data suggests that disrupting the retrograde motor leads to a disruption in retrograde neurofilament transport and to the inactivation of the

114 anterograde motor, leading to the reciprocal disruption of anterograde transport.

The increase in anterograde neurofilament transport observed in SCG neurons could offer us a mechanistic insight into the coordination and cross-talk between the oppositely-directed motors, and further research should focus on this phenomenon.

115

Chapter 5: Discussion

5.1. Role of myosin Va in short-range transport

Our measurement of neurofilament transport frequency along axons of SCG neurons from dilute lethal mice reveals that in the absence of myosin Va neurofilaments are transported less efficiently than in wild type neurons. The contribution of myosin Va in the movement appears to be subtle, as evident from the observed 27% decrease in frequency. Yet, this subtle contribution was expected. The general conception regarding actin-dependent transport has always been that cargo associated with actin filaments are translocated for short distances to position transported structures within discrete subcompartments along the axons. The long-range transport, on the other hand, is usually reserved for microtubule-based transport (Langford, 2002; Brown et al., 2004). A previous study from the lab of Mark Black investigated the role of actin-based transport of neurofilaments (Francis et al., 2005). The authors observed 40% reduction in transport frequencies after actin depletion using latrunculin, while depolymerization of axonal microtubules completely inhibited neurofilament transport. The study suggests that actin-based transport is not required for

116 neurofilament movement along the axons, although it could contribute in a subtle manner.

Moreover, measurement of the transport kinetics in our experiments indicates that the absence of myosin Va from the axonal compartment has no effect on the efficiency at which the moving filaments are transported along their tracks. In other words, if a neurofilament is moving, it moves with the same speed and for comparable distances whether myosin Va is present or absent from the axons.

Trivedi et al. (2007) have previously shown using computational and kinetic analyses that only ~3% of the total neurofilament axonal population is moving at any point in time (“on-track”). The other 97% of the neurofilaments are stuck in a prolonged pausing state (“off-track”). Experiments from the study were performed in rat SCG neurons, transfected with PAGFP-NFM, and the neurofilament movement was monitored after photoactivation. The power of this technique is that it allows the observation, for prolonged periods of time, of the

“off-track” population of neurofilaments. Therefore, our observations of neurofilament movement in mouse SCG neurons through naturally occurring gaps could possibly under-estimate the role of myosin Va on neurofilament behavior as it restricts us to monitoring the movement of 3% of the total neurofilament population, overlooking the effect on the “off-track” polymers.

117 For this reason, we utilized pulse-escape fluorescence photoactivation technique in neurons from dilute lethal and wild type mice. These experiments revealed a significant increase in neurofilament pausing in DRG neurons from DLS/LeJ and dl20J mice as compared to wild type. Mean time for departure of neurofilaments was 466 and 1294 minutes in DLS/LeJ and dl20J, respectively. This means that, on average, neurofilaments from DLS/LeJ and dl20J mice pause for 2.5 and 13.5 hours more than their counterparts from wild type mice, respectively. Thus, on long time scales, the difference in the transport kinetics is extremely significant.

In the previous study (Trivedi et al., 2007), the mean time for departure of neurofilaments from rat SCG neurons was 96 minutes, which is considerably shorter than in mouse DRG neurons with mean departure times of 315 and 482 minutes from the DLS/LeJ and dl20J experiment, respectively. As neurofilaments pause for longer periods in DRG neurons, their residence time in the axon increases, which could possibly explain the higher neurofilament content of DRG compared to SCG neurons.

The significant increase in neurofilament pausing in the absence of myosin Va indicates a possible role for the motor protein in facilitating neurofilament movement in the axons. One possible explanation for the increase in neurofilament pausing is a change in the phosphorylation state of neurofilaments in the absence of myosin Va, specifically the phosphorylation of KSP motifs in the tail domains of NFM and NFH (for details, refer to section 1.2.3). To examine

118 this possibility, we performed quantitative Western blotting in wild type and dilute lethal spinal cord using three phospho-specific antibodies. One such antibody,

RT97, has been extensively characterized. The RT97 epitope is generated by phosphorylation at KSPXK and KSPXXXK sites in the tail domain of NFH

(Veeranna et al., 2008), and phosphorylation at RT97 epitopes correlates directly with the slowing of neurofilament transport in mouse optic nerve (Ackerley et al.,

2000; Sanchez et al., 2000). However, we found no difference in neurofilament phosphorylation using this Ab or either of the other phospho-dependent Abs.

This suggests that altered phosphorylation is not the explanation for the increased pausing of neurofilaments in the absence of myosin Va.

These data do not completely rule out the possibility that myosin Va has an indirect effect on neurofilament movement, but some evidence suggests that myosin Va may be a motor for neurofilaments. Rao et al. (2002) have demonstrated that myosin Va interacts with NFL. Moreover, myosin Va is considered a short-range motor for many of the axonal organelles especially in regions devoid of microtubule tracks, and where actin is more abundant.

However, recent studies have proposed that actin filaments can also provide support for axonal transport in the central region of the axon (Bearer and Reese,

1999).

119 Actin filaments as tracks for axonal transport:

Ultrastructural studies in neuronal axons have demonstrated that actin filaments are abundant at the cortical region adjacent to the axonal plasma membrane.

The microfilaments in the subaxolemmal region of the squid giant axon were observed by thin section and freeze-etch replica electron microscopy. These microfilaments were associated with the axolemma. Microtubules ran parallel to the axolemma and embedded in the meshwork of the microfilaments (Figure 5.1;

(Metuzals and Tasaki, 1978; Hodge and Adelman, 1980). Fluorescent probes for actin stained the axonal cortex more intensely than the central region of the axons, which was used as evidence that microfilaments are restricted to the subaxolemmal environment (Spooner and Holladay, 1981; Spooner et al., 1982).

Some studies utilized extraction after fixation, and compared actin organization in the growth cone to those in the neurites of chick sensory neurons to deduce that filamentous actin is concentrated in the growth cones and is infrequent in neurites (Letourneau, 1983).

120

Figure 5.1. A schematic of the architecture of the subaxolemmal cytoskeleton of a squid giant axon Longitudinally-oriented microtubules (MT) are embedded in a fine meshwork containing microfilaments (AF) and other proteins. PL: plasmalemma. Such a model allows for exchanging cargo between the microtubule and microfilament tracks. (Tsukita et al., 1986).

This view contradicted the observations of Ross et al. (Ross et al., 1975) and

Kuczmarski and Rosenbaum (Kuczmarski and Rosenbaum, 1979) who used whole mount transmission electron microscopy to look at microfilament organization in the neurites of cultured DRG neurons and reported the presence of myosin-associated microfilaments in the neurites and their spatial association with microtubules. Quantitative biochemical analyses also showed that actin filaments were abundant in the central axonal cytoskeleton, and associated with microtubules (Morris and Lasek, 1984; Heriot et al., 1985; Fath and Lasek,

1988).

Injection of , which severs actin filaments, arrests organelle movement suggesting that microtubules by themselves are not sufficient to support axonal

121

transport (Brady et al., 1984). The idea that cargo can move along both actin and microtubules in neurons, therefore, is not a novel idea. In hippocampal neurons, mitochondria can move on both actin filaments and microtubules

(Morris and Hollenbeck, 1995) and organelles from neuronal axons were isolated in association with myosin motors (Bearer et al., 1993; Bearer et al., 1996b;

Bearer et al., 1996a).

Figure 5.2. A schematic showing actin filaments as short-distance tracks for organelle transport Bearer and Reese (1999) suggest that actin tracks within microtubule bundles can mediate transfer of organelles from one microtubule to the next. In this model organelles move for short distances on actin filaments.

Using confocal and electron microscopy, Bearer and Reese followed those observations with a study showing that longitudinal actin filaments are abundant in neuronal axons and are interwoven among the long microtubule tracks (Bearer and Reese, 1999). The study also suggests bidirectional transport along the actin tracks, as deduced from the jumbled polarity of the filaments.

122

Microtubules are discontinuous along the axons, so actin filaments could possibly act as bridges between microtubule tracks. In this way, myosin Va could facilitate the transition of neurofilaments from one microtubule to another.

Alternatively, myosin could move cargo laterally on actin tracks for short distances between parallel microtubules (Figure 5.3). It is tempting to speculate that lateral movement gains more importance in fat axons, where neurofilaments generally outnumber microtubules, and thus, many neurofilaments may be at a distance from the nearest microtubule track. By transporting neurofilaments along actin filaments in the radial dimension of the axon, myosin Va may increase the frequency with which these stranded neurofilaments encounter microtubule tracks, thereby decreasing the duration of the off-track pauses.

Figure 5.3. Myosin Va as a short-distance motor for neurofilament movement We speculate that myosin Va is capable of moving neurofilaments longitudinally or radially for short distances along actin filaments in axons. According to this hypothesis, myosin Va decreases the duration of the prolonged off-track pauses by delivering off-track neurofilaments to their microtubule tracks. In the absence of myosin Va, off-track neurofilaments may become stranded away from their microtubule tracks for longer periods of time, and on-track neurofilaments that disengage from their tracks may take longer to re- engage. This would lead to neurofilament accumulation along the axons, which could explain the increase in neurofilament number in axons of dilute lethal mice reported by Nixon and colleagues.

123 5.2. The role of dynein/dynactin in neurofilament transport

We investigated the role of the dynein/dynactin complex in neurofilament transport using a series of manipulations targeting different subunits of the complex in SCG neurons. Results obtained from each manipulation were assessed in context of its effects on retrograde transport, anterograde transport, and movement kinetics when possible. One experiment was performed on cortical neurons with varying transfection concentrations at two different time points to investigate the possibility of cell-type, expression level, and temporal dependence.

Cytoplasmic dynein is a complex molecule composed of two heavy chains, multiple intermediate, light intermediate and light chains. The two heavy chains fold to form globular heads on flexible stalks that dimerize at their ends. Each of the globular heads forms a motor domain where ATP binding and hydrolysis takes place. We used RNA interference to knock down dynein heavy chain, and inhibit dynein activity. As expected, this caused a significant decrease in retrograde neurofilament transport. Surprisingly, we also observed a reciprocal effect on anterograde transport. This indicates that the retrograde motor is somehow involved in anterograde transport. To confirm this result, we used a function blocking monoclonal antibody directed against dynein intermediate chain, which is important for dynein biding to cargo and dynactin (Karki and

Holzbaur, 1995; Vaughan and Vallee, 1995; Steffen et al., 1996). The expression and regulation of dynein intermediate chain is cell and tissue specific,

124 and developmentally regulated (Dillman et al., 1996; Pfister et al., 1996a; Pfister et al., 1996b). Vaughan and Vallee have identified two dynein intermediate chains genes and five splice variants in rat brain (Vaughan and Vallee, 1995), while the monoclonal antibody used in our study is capable of binding to at least six different intermediate chain isoforms (Brill and Pfister, 2000). This treatment in SCG neurons resulted in dramatic impairment of retrograde movement, which was mirrored in the anterograde direction.

In vivo, the activity of dynein requires dynactin, which increases the processivity of the motor by binding to microtubule tracks (King and Schroer, 2000a).

Dynactin is also a large protein complex, composed of 11 subunits (for details refer to the introduction section 1.4). Overexpression of dynamitin/p50- a dynactin subunit- disrupts dynein function by disrupting the dynactin complex

(Echeverri et al., 1996; Melkonian et al., 2007). To confirm the above results using a dominant negative strategy, we transfected SCG neurons with chicken dynamitin. Once again, we observed an inhibition of retrograde transport, accompanied by an impairment of anterograde movements.

p150Glued is the largest subunit in the dynactin complex and acts as a binding partner of both microtubules and dynein (Karki and Holzbaur, 1995; Vaughan and Vallee, 1995; Waterman-Storer et al., 1995; Quintne et al., 1999; Vaughan et al., 2001). We transfected cortical neurons with DsRed-tagged CC1 domain of p150, which is the domain responsible for binding the dynein intermediate chain

125 (Quintyne et al., 1999). Previously, it was reported that overexpression of this construct inhibits dynein activity in a dominant negative manner by dissociating dynein from dynactin without affecting the integrity of the two complexes

(Quintyne and Schroer, 2002). This experiment confirmed our previous findings as it resulted in an impairment of retrograde transport that was mirrored by the anterograde direction using two different concentrations of the construct and at two different time points after plating.

Microtubule tracks have a uniform polarity along neuronal axons, with their plus ends pointing towards axonal tips (Heidemann et al., 1981). Consequently, the minus end-directed dynein/dynactin motor could not function as a retrograde motor, but it can be involved in activating kinesin-dependent transport. This idea that the two motors are interdependent has been proposed earlier to explain transport dynamics of membranous organelles, vesicles and other cargo. For example, monoclonal antibodies to kinesin-1 or dynactin inhibited both anterograde and retrograde organelle movement in squid axoplasm (Brady et al.,

1990). Mutations in kinesin-1 or dynein/dynactin exhibited general defects in both directions of transport in Drosophila larval segmented nerve (Hurd and

Saxton, 1996; Martin et al., 1999), and severely impaired movements of lipid droplets in Drosophila embryos (Gross et al., 2002). Knocking down kinesin-1 or dynein inhibited the movement of ribonucleoprotein granules in Drosophila S2 cells, and dynamitin or CC1-p150 overexpression causes an inhibition of

126 bidirectional transport of dense-core vesicles in the axons and dendrites of cultured hippocampal neurons (Kwinter et al., 2009).

Figure 5.4. Three possible models for bidirectional transport In the ‘tug-of-war’ model (a), both antero- and retrograde motors are simultaneously engaged. Cargo move in the direction of motors exerting more force. In the ‘exclusionary presence’ model (b), only one set of motors can engage the cargo at any given time. In the ‘coordination’ model (c), both sets of motors are on the cargo simultaneously, but the activity of the motors is coordinated so that they do not interfere with each other’s activity. (Gross, 2004).

One of the main questions in the field of axonal transport at the moment is regulation of bidirectional transport and the mechanisms governing the nature of interaction between the two oppositely directed motors to control transport.

Different models have been proposed to this effect, each of which has its set of assumptions and expectations (Gross, 2004). Current research is focused on understanding the relationship between dynein and kinesins in the context of the predictions from these models, which can be summarized as follows:

5.2.1. A simple tug-of-war model

According to this model, axonal cargo is bound by two fully functional and active oppositely-directed motors. Each motor is pulling to move in its direction, and the

127 cargo is caught in a tug-of-war situation. In such a system, the motor that exerts more force to win this tug-of-war decides the net direction of transport. For example, for a neurofilament to move in the anterograde direction, the number of active kinesin motors should increase, or the number of dynein should decrease.

In this case, if we were to impair the retrograde motor, we would expect an increase in transport frequency in the opposite direction. A decrease in the stall forces acting on active kinesins would also mean an increase in anterograde velocity. Our experimental results mentioned earlier do not show an increase in anterograde frequency. We also examined the transport kinetics in the anterograde and retrograde directions from SCG neurons transfected with dynamitin, and observed no change in velocity as compared to wild type.

Therefore, our data do not support a simple tug-of-war system.

5.2.2. The exclusionary presence model

The idea behind this model is simple: once a cargo is bound by dynein, it cannot be associated with kinesin, and vice versa. This means that the two directions of transport are completely independent from each other, and manipulation of one motor should have no effect on the other’s activity. One might argue that the absence of one motor could make the cargo more available for the other, but even in such a case, there should be no effect on the velocity of transport.

Studies have revealed that cargo in rat brain and in vitro associated with dynein and kinesin, arguing against the exclusionary model’s prediction, and fluorescently-tagged dynein intermediate chain was observed to translocate

128 bidirectionally as it was associated with cargo, indicating that it is within a system which includes kinesin motors (Rogers et al., 1997; Murray et al., 2000; Ligon et al., 2004). Our results clearly indicate that disrupting retrograde transport also impairs anterograde transport, with no change in movement kinetics, which does not agree with the exclusionary model’s predictions.

5.2.3. The coordination model

This model hypothesizes that anterograde and retrograde motors work together in a well-coordinated system, and the motors’ activity is tightly regulated such that when one motor is active, the other motor is turned off. This coordination could either take place through a direct physical interaction, or through a third molecule that could mediate the cross-talk between the two motors. For example

Klar, a product of a previously unstudied gene, has been proposed as a regulator of dynein-kinesin interaction in Drosophila early embryos (Welte et al., 1998). In embryos with mutant klar, the switch that synchronizes the activity of the motors is absent, resulting in a tug-of-war-type of motion, where stalling forces equilibrate, and vesicle transport is impaired (Welte et al., 1998). Similar roles have been attributed to Halo and LSD2 (Welte et al., 2005).

Alternatively, motors could be regulated by mechanical coupling: a motor disengages from the cargo or is inactive under a certain amount of load, which could simply be the pull in the opposite direction. Hence, the coordination model suggests that both motors can bind cargo simultaneously, and inhibition of one

129 does not alter the velocity of transport in the other direction. Such an assumption could easily fit the data obtained in our study, where transport is inhibited upon disruption of the retrograde motor, while the kinetics of transport remain unchanged.

5.2.4. The coordinated tug-of-war model

A novel model for bidirectional transport was proposed by Muller et al. who theoretically studied the mechanism of bidirectional transport (Müller et al.,

2008). Using transport properties of individual motors as measured in single- molecule experiments, they suggest that a tug-of-war model could be very highly cooperative, exhibiting seven different “motility regimes” or states of transport.

These seven possibilities are: fast anterograde transport, fast retrograde transport, pausing, anterograde to retrograde reversals without pause, and vice versa, and retrograde to anterograde reversals with pause, and vice versa.

Using this theoretical model, and by modifying single-motor properties, the authors tried to mimic the effects of motor mutations or regulatory processes, and found cases where motion was affected in only one direction, cases where motion was impaired in one direction and enhanced by the other, and cases where motion was enhanced or impaired in both directions. This variation agrees with our experimental variations, and with the discrepancy between our dynein heavy chain knock-down data and the observation of He et al. (2005). The model has not been closely tested in vitro or in vivo, and applies the properties of single motors to a mathematical model to arrive at its conclusions, but it could

130 provide an explanation for the contradicting observations we have encountered using the same expression construct in different neuronal cell types.

5.2.5. Possible interpretations of CC1-p150 overexpression data

The only observation that does not fit the prediction of a coordination model in our results comes from SCG neurons overexpressing the CC1 fragment of p150.

The same fragment was overexpressed in cortical neurons and caused an inhibition of bidirectional transport of neurofilaments at two different time points and using two different concentrations. In SCG neurons, though, we observed a disruption in retrograde transport, but an increase in anterograde frequency. The kinetics of transport in the anterograde direction remained unchanged as compared to wild type. The reason for the difference in the effect is not clear so far. CC1-p150 overexpression dissociates dynein from dynactin and keeps both complexes intact (Quintyne and Schroer, 2002). It was proposed earlier that dynactin is required for transport activity of dynein and kinesin II in Xenopus melanophores (Deacon et al., 2003). It is not known if an intact dynamitin would activate kinesin-1, which is thought to be responsible for neurofilament anterograde transport (for details refer to introduction section 1.3). But a possible explanation for our data is that the impairment of retrograde transport is due to the dissociation of the dynein/dynactin complex, while the anterograde motor remains active, as its activator -dynactin- is available. As dynactin acts like a switch that synchronizes anterograde and retrograde motors, anterograde frequency would be expected to increase because kinesin is not “turned off” in

131 the absence of retrograde motor activity. The kinetics of transport in the anterograde direction would thus remain unchanged, as observed in our data.

Yet, in cortical neurons, this does not seem to be the case, possibly because different kinesins are involved in anterograde transport, but we have not investigated this possibility thus far. In fact, there is precedent for the differential effect of dynactin manipulation in different cell types: In Xenopus melanophores, overexpressing CC1-p150 inhibits both directions of melanosome transport

(Deacon et al., 2003). On the other hand, overexpressing p50/dynamitin in HeLa and TC7 cells causes an impairment of retrograde and enhancement of anterograde transport frequency of adenovirus particles (Suomalainen et al.,

1999).

5.3. Future directions

We plan to address the discrepancy between our result after DHC siRNA injections and the result obtained by He et al. (2005). In that study, the authors used rat SCG neurons, and different siRNA sequences that resulted in a more complete knock down on DHC. Moreover, the siRNA duplexes were introduced to the cells using electroporation before plating, while our duplexes were injected into the cells after extending axons with neurofilaments along their length 2 days after plating. How these differences account for different effects on neurofilament transport is not clear. We could possibly address the question by introducing our methodology to rat SCG neurons, or by electroporation of the

132 siRNA construct into mouse SCG neurons, but this could risk the viability of the cells, as we know from previous experience.

We propose here that myosin Va is a motor for neurofilament transport, and the association of NFL and the motor has been already demonstrated. To confirm that myosin is directly involved in the transport of neurofilaments on actin tracks, we can try to rescue the motor activity in the neurons by direct injection of the protein into the neuronal cytoplasm. Moreover, we can use actin-depolymerizing agents such as latrunculin to mimic the effect of myosin Va disruption on neurofilament transport. This would confirm that the effects observed in the absence of myosin Va are a result of the inability of neurofilaments to translocate along actin tracks.

To look further into the regulation of long-range neurofilament transport, it would be interesting to isolate neurofilament polymers associated with both dynein/dynactin and kinesin motors. Previous studies have co-purified dynein/dynactin with neurofilaments (Shah et al., 2000), but nothing is known about the association with kinesins. If it were indeed the case that neurofilaments are bound by a motor protein complex, mechanistic insight can be gained by using dominant negative treatments to dissociate the motors from the cargo. For example, if the overexpression of CC1-p150 fragment were to dissociate dynein from neurofilaments without releasing kinesins, it would explain the increase in anterograde frequency we observe in SCG neurons. We then

133 would expect that the same treatment in cortical neurons would release the whole kinesin/dynein complex from neurofilaments.

Investigating a possible explanation of our results by the coordinated tug-of-war model proposed by Muller et al. (2008) could prove very useful. To do so, it would be necessary to utilize direct measurements of stall forces under selective impairment of either dynein or kinesin. Changes in directionality of transport as motors bind/dissociate from neurofilaments, and the changes in stall forces as these events occur is not currently possible for neurofilament polymers, although it could be applied on individual lipid droplets from Drosophila embryos where the use of optical tweezers and estimation of the number of motors bound to cargo is possible using direct measurements (Shubeita et al., 2008). In their study,

Shubeita et al. argue that the number of motor proteins bound to the cargo has no effect on the run lengths and kinetics of transport of lipid droplets, but the much longer and larger neurofilament polymers could present a different case, where the number of motors bound to the polymer could be a factor in the net directionality and the run length of movement. Variations in the expression levels of our dominant negative treatments could therefore have different effects on the directionality of transport as the numbers of available motors change. Our current use of transiently-expressing cells does not allow a tight control over the expression levels, and the use of a stably-expressing cell-line could help overcome this limitation.

134 Huang et al. were able to demonstrate that myosin Va can directly interact with mouse ubiquitous kinesin heavy chain, prompting them to propose that the transfer of vesicles from microtubules to actin and vice versa is possible (Huang et al., 1999). Probing for this complex on neurofilament polymers along the axons in association with dynein/dynactin and manipulating its binding properties using different treatments could be a major step in formulating a general model for neurofilament axonal transport.

5.4. A general model for neurofilament axonal transport

We have investigated the bidirectional transport of neurofilament polymers along neuronal axons using live-cell imaging techniques in a variety of neuronal cell types. Our study puts forward a model for neurofilament transport that encompasses the contributions of three different molecular motors and two cytoskeletal tracks along which short- and long-range transport takes place

(Figure 5.5).

We propose that myosin Va is a short-range motor responsible for the lateral translocation of neurofilaments on short actin tracks, rendering them available for long-range transport by delivering them to microtubules or facilitating their association with microtubules. This actin-based movement could be responsible for the transfer of “off-track” neurofilaments stranded away from the microtubules to “on-track” neurofilaments readily available for microtubule-based transport. In

135 the absence of myosin Va activity, neurofilaments are not capable of moving from one microtubule to the other, or switching from the “off-track” state to “on- track”, causing an axonal accumulation of neurofilaments, and an increase in their numbers along the axons as reported by Rao et al. (2002).

When neurofilaments are in the vicinity of microtubules, they are available for long-range transport, which is carried out by two oppositely directed motors: dynein/dynactin and kinesins. Here, we propose that dynein/dynactin and kinesins are functionally interdependent. Using both, an acute (2-6 hours after

Ab injections) and a chronic (3-7 days post dominant negative construct transfections and 6 days post-siRNA injections) inhibition of the retrograde motor, we report impairment of retrograde transport, which, surprisingly, was mirrored by anterograde transport, except in one manipulation. Previous results from our lab have also demonstrated, using kinesin knock-out and dominant negative treatments, that disruption of the anterograde motor can also inhibit retrograde transport, confirming our hypothesis that kinesins and dynein are functionally coupled. Therefore, the movement of neurofilaments along neuronal axons is a collaborative effort of opposite-polarity motors that move on microtubules and actin filaments, an idea that is supported by a report showing a direct interaction between actin- and microtubule-based motors (Huang et al., 1999). Disrupting the activity of any of these motors or the tracks they use for movement can cause an impairment in neurofilament translocation in the cells, resulting in

136 accumulations of the protein, and possible blockage of axonal transport that is considered an early pathological marker of many neurodegenerative diseases.

Figure 5.5. A unified model for neurofilament axonal transport The majority of axonal neurofilaments are “off-track” (~97%), with pauses extending for several hours (Trivedi et al., 2007). We propose that the “off-track” filaments are stranded away from microtubules. Myosin Va can decrease the duration of prolonged off-track pauses by delivering off-track neurofilaments to microtubules, rendering them available for kinesin and dynein activity. These “on-track” neurofilaments (~3% of the total neurofilament population) move along the axons bidirectionally under the coordinated activity of both sets of motors. The inhibition of any one of the actin-dependent or microtubule-dependent motors could lead to the disruption of neurofilament axonal transport.

137

REFERENCES

Ackerley S, Grierson AJ, Brownlees J, Thornhill P, Anderton BH, Leigh PN, Shaw CE, Miller CC (2000) Glutamate slows axonal transport of neurofilaments in transfected neurons. In: J Cell Biol, pp 165-176.

Ackerley S, Thornhill P, Grierson AJ, Brownlees J, Anderton BH, Leigh PN, Shaw CE, Miller CC (2003) Neurofilament heavy chain side arm phosphorylation regulates axonal transport of neurofilaments. J Cell Biol 161:489-495.

Aebi U, Cohn J, Buhle L, Gerace L (1986) The nuclear lamina is a meshwork of intermediate-type filaments. In: Nature, pp 560-564.

Aizawa H, Sekine Y, Takemura R, Zhang Z, Nangaku M, Hirokawa N (1992) Kinesin family in murine central nervous system. J Cell Biol 119:1287- 1296.

Allan V (1996) Motor proteins: a dynamic duo. Curr Biol 6:630-633.

Angelides KJ, Smith KE, Takeda M (1989) Assembly and exchange of intermediate filament proteins of neurons: neurofilaments are dynamic structures. In: J Cell Biol, pp 1495-1506.

Archer DR, Watson DF, Griffin JW (1994) Phosphorylation-dependent immunoreactivity of neurofilaments and the rate of slow axonal transport in the central and peripheral axons of the rat dorsal root ganglion. In: J Neurochem, pp 1119-1125.

Asbury AK, Gale MK, Cox SC, Baringer JR, Berg BO (1972) Giant axonal neuropathy--a unique case with segmental neurofilamentous masses. Acta Neuropathol 20:237-247.

Averback P (1981) Unusual particles in motor neuron disease. Arch Pathol Lab Med 105:490-493.

Bearer EL, Reese TS (1999) Association of actin filaments with axonal microtubule tracts. In: J Neurocytol, pp 85-98.

138 Bearer EL, DeGiorgis JA, Medeiros NA, Reese TS (1996a) Actin-based motility of isolated axoplasmic organelles. In: Cell Motil Cytoskeleton, pp 106-114.

Bearer EL, DeGiorgis JA, Bodner RA, Kao AW, Reese TS (1993) Evidence for myosin motors on organelles in squid axoplasm. In: Proc Natl Acad Sci USA, pp 11252-11256.

Bearer EL, DeGiorgis JA, Jaffe H, Medeiros NA, Reese TS (1996b) An axoplasmic myosin with a calmodulin-like light chain. In: Proc Natl Acad Sci USA, pp 6064-6068.

Beaulieu JM, Nguyen MD, Julien JP (1999) Late onset of motor neurons in mice overexpressing wild-type peripherin. In: J Cell Biol, pp 531-544.

Berg BO, Rosenberg SH, Asbury AK (1972) Giant axonal neuropathy. Pediatrics 49:894-899.

Berg JS, Powell BC, Cheney RE (2001) A millennial myosin census. In: Mol Biol Cell, pp 780-794.

Blaker WD, Goodrum JF, Morell P (1981) Axonal transport of the mitochondria- specific lipid, diphosphatidylglycerol, in the rat visual system. J Cell Biol 89:579-584.

Bomont P, Cavalier L, Blondeau F, Ben Hamida C, Belal S, Tazir M, Demir E, Topaloglu H, Korinthenberg R, Tuysuz B, Landrieu P, Hentati F, Koenig M (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 DR, Wong PC, Becher MW, Pardo CA, Lee MK, Xu ZS, Thinakaran G, Jenkins NA, Copeland NG, Sisodia SS, Cleveland DW, Price DL, Hoffman PN (1998) Axonal transport of mutant superoxide dismutase 1 and focal axonal abnormalities in the proximal axons of transgenic mice. In: Neurobiol Dis, pp 27-35.

Brady ST (1985) A novel brain ATPase with properties expected for the fast axonal transport motor. Nature 317:73-75.

Brady ST, Pfister KK, Bloom GS (1990) A monoclonal antibody against kinesin inhibits both anterograde and retrograde fast axonal transport in squid axoplasm. In: Proc Natl Acad Sci USA, pp 1061-1065.

Brady ST, Lasek RJ, Allen RD, Yin HL, Stossel TP (1984) Gelsolin inhibition of fast axonal transport indicates a requirement for actin microfilaments. In: Nature, pp 56-58.

139 Bridgman PC (1999) Myosin Va movements in normal and dilute-lethal axons provide support for a dual filament motor complex. In: J Cell Biol, pp 1045- 1060.

Bridgman PC (2004) Myosin-dependent transport in neurons. In: J Neurobiol, pp 164-174.

Brill LB, Pfister KK (2000) Biochemical and molecular analysis of the mammalian cytoplasmic dynein intermediate chain. In: Methods, pp 307-316.

Brown A (2000) Slow axonal transport: stop and go traffic in the axon. In: Nat Rev Mol Cell Biol, pp 153-156.

Brown A (2003) Axonal transport of membranous and nonmembranous cargoes: a unified perspective. In: J Cell Biol, pp 817-821.

Brown A (2008) Slow Axonal Transport. In: Elsevier Encyclopedia, p 9.

Brownlees J, Ackerley S, Grierson AJ, Jacobsen NJ, Shea K, Anderton BH, Leigh PN, Shaw CE, Miller CC (2002) Charcot-Marie-Tooth disease neurofilament mutations disrupt neurofilament assembly and axonal transport. Hum Mol Genet 11:2837-2844.

Burkhardt JK, Echeverri CJ, Nilsson T, Vallee RB (1997) Overexpression of the dynamitin (p50) subunit of the dynactin complex disrupts dynein- dependent maintenance of membrane organelle distribution. In: J Cell Biol, pp 469-484.

Cairns NJ, Perry RH, Jaros E, Burn D, McKeith IG, Lowe JS, Holton J, Rossor MN, Skullerud K, Duyckaerts C, Cruz-Sanchez FF, Lantos PL (2003) Patients with a novel neurofilamentopathy: dementia with neurofilament inclusions. Neurosci Lett 341:177-180.

Carden MJ, Schlaepfer WW, Lee VM (1985) The structure, biochemical properties, and immunogenicity of neurofilament peripheral regions are determined by phosphorylation state. In: J Biol Chem, pp 9805-9817.

Carpenter S (1968) Proximal axonal enlargement in motor neuron disease. Neurology 18:841-851.

Carter J, Gragerov A, Konvicka K, Elder GA, Weinstein H, Lazzarini RA (1998) Neurofilament (NF) assembly; divergent characteristics of human and rodent NF-L subunits. In: J Biol Chem, pp 5101-5108.

Chan WK, Yabe JT, Pimenta AF, Ortiz D, Shea TB (2003) Growth cones contain a dynamic population of neurofilament subunits. Cell Motil Cytoskeleton 54:195-207.

140 Chen J, Nakata T, Zhang Z, Hirokawa N (2000) The C-terminal tail domain of neurofilament protein-H (NF-H) forms the crossbridges and regulates neurofilament bundle formation. J Cell Sci 113 Pt 21:3861-3869.

Cheney RE, O'Shea MK, Heuser JE, Coelho MV, Wolenski JS, Espreafico EM, Forscher P, Larson RE, Mooseker MS (1993) Brain myosin-V is a two- headed unconventional myosin with motor activity. Cell 75:13-23.

Chiu FC, Barnes EA, Das K, Haley J, Socolow P, Macaluso FP, Fant J (1989) Characterization of a novel 66 kd subunit of mammalian neurofilaments. Neuron 2:1435-1445.

Cochard P, Paulin D (1984) Initial expression of neurofilaments and vimentin in the central and peripheral nervous system of the mouse embryo in vivo. J Neurosci 4:2080-2094.

Cote F, Collard JF, Julien JP (1993) Progressive neuronopathy in transgenic mice expressing the human neurofilament heavy gene: a mouse model of amyotrophic lateral sclerosis. Cell 73:35-46.

Couillard-Després S, Zhu Q, Wong PC, Price DL, Cleveland DW, Julien JP (1998) Protective effect of neurofilament heavy gene overexpression in motor neuron disease induced by mutant superoxide dismutase. In: Proc Natl Acad Sci USA, pp 9626-9630.

Dahlstrand J, Collins VP, Lendahl U (1992a) Expression of the class VI intermediate filament nestin in human central nervous system tumors. Cancer Res 52:5334-5341.

Dahlstrand J, Zimmerman LB, McKay RD, Lendahl U (1992b) Characterization of the human nestin gene reveals a close evolutionary relationship to neurofilaments. J Cell Sci 103 ( Pt 2):589-597.

De Jonghe P, Mersivanova I, Nelis E, Del Favero J, Martin JJ, Van Broeckhoven C, Evgrafov O, Timmerman V (2001) Further evidence that neurofilament light chain gene mutations can cause Charcot-Marie-Tooth disease type 2E. Ann Neurol 49:245-249.

Deacon SW, Serpinskaya AS, Vaughan PS, Lopez Fanarraga M, Vernos I, Vaughan KT, Gelfand VI (2003) Dynactin is required for bidirectional organelle transport. J Cell Biol 160:297-301.

Delisle MB, Carpenter S (1984) Neurofibrillary axonal swellings and amyotrophic lateral sclerosis. J Neurol Sci 63:241-250.

Diefenbach RJ, Mackay JP, Armati PJ, Cunningham AL (1998) The C-terminal region of the stalk domain of ubiquitous human kinesin heavy chain

141 contains the binding site for kinesin light chain. Biochemistry 37:16663- 16670.

Diggle P, Heagerty P, Liang K-Y, Zeger S (2002) Analysis of longitudinal data, 2 Edition. New York: Oxford UP.

Dillman JF, Pfister KK (1994) Differential phosphorylation in vivo of cytoplasmic dynein associated with anterogradely moving organelles. In: J Cell Biol, pp 1671-1681.

Dillman JF, 3rd, Dabney LP, Karki S, Paschal BM, Holzbaur EL, Pfister KK (1996) Functional analysis of dynactin and cytoplasmic dynein in slow axonal transport. J Neurosci 16:6742-6752.

Drager UC (1983) Coexistence of neurofilaments and vimentin in a neurone of adult mouse retina. Nature 303:169-172.

Echeverri CJ, Paschal BM, Vaughan KT, Vallee RB (1996) Molecular characterization of the 50-kD subunit of dynactin reveals function for the complex in chromosome alignment and spindle organization during mitosis. In: J Cell Biol, pp 617-633.

Eckley DM, Gill SR, Melkonian KA, Bingham JB, Goodson HV, Heuser JE, Schroer TA (1999) Analysis of dynactin subcomplexes reveals a novel actin-related protein associated with the arp1 minifilament pointed end. J Cell Biol 147:307-320.

Elder GA, Friedrich VL, Bosco P, Kang C, Gourov A, Tu PH, Lee VM, Lazzarini RA (1998a) Absence of the mid-sized neurofilament subunit decreases axonal calibers, levels of light neurofilament (NF-L), and neurofilament content. In: J Cell Biol, pp 727-739.

Elder GA, Friedrich VL, Kang C, Bosco P, Gourov A, Tu PH, Zhang B, Lee VM, Lazzarini RA (1998b) Requirement of heavy neurofilament subunit in the development of axons with large calibers. In: J Cell Biol, pp 195-205.

Elhanany E, Jaffe H, Link WT, Sheeley DM, Gainer H, Pant HC (1994) Identification of endogenously phosphorylated KSP sites in the high- molecular-weight rat neurofilament protein. J Neurochem 63:2324-2335.

Endow SA, Waligora KW (1998) Determinants of kinesin motor polarity. Science 281:1200-1202.

Evan GI, Lewis GK, Ramsay G, Bishop JM (1985) Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product. Mol Cell Biol 5:3610-3616.

142 Evans LL, Hammer J, Bridgman PC (1997) Subcellular localization of myosin V in nerve growth cones and outgrowth from dilute-lethal neurons. J Cell Sci 110 ( Pt 4):439-449.

Eyer J, Peterson A (1994) Neurofilament-deficient axons and perikaryal aggregates in viable transgenic mice expressing a neurofilament-beta- galactosidase fusion protein. Neuron 12:389-405.

Eyer J, Cleveland DW, Wong PC, Peterson AC (1998) Pathogenesis of two axonopathies does not require axonal neurofilaments. Nature 391:584- 587.

Fabrizi GM, Cavallaro T, Angiari C, Bertolasi L, Cabrini I, Ferrarini M, Rizzuto N (2004) Giant axon and neurofilament accumulation in Charcot-Marie- Tooth disease type 2E. Neurology 62:1429-1431.

Fabrizi GM, Cavallaro T, Angiari C, Cabrini I, Taioli F, Malerba G, Bertolasi L, Rizzuto N (2007) Charcot-Marie-Tooth disease type 2E, a disorder of the cytoskeleton. Brain 130:394-403.

Fath KR, Lasek RJ (1988) Two classes of actin microfilaments are associated with the inner cytoskeleton of axons. In: J Cell Biol, pp 613-621.

Fois A, Balestri P, Farnetani MA, Berardi R, Mattei R, Laurenzi E, Alessandrini C, Gerli R, Ribuffo A, Calvieri S (1985) Giant axonal neuropathy. Endocrinological and histological studies. Eur J Pediatr 144:274-280.

Forman MS, Trojanowski JQ, Lee VM (2004) Neurodegenerative diseases: a decade of discoveries paves the way for therapeutic breakthroughs. Nat Med 10:1055-1063.

Friede RL, Samorajski T (1970) Axon caliber related to neurofilaments and microtubules in sciatic nerve fibers of rats and mice. Anat Rec 167:379- 387.

Galloway PG, Mulvihill P, Perry G (1992) Filaments of Lewy bodies contain insoluble cytoskeletal elements. Am J Pathol 140:809-822.

Garcia ML, Lobsiger CS, Shah SB, Deerinck TJ, Crum J, Young D, Ward CM, Crawford TO, Gotow T, Uchiyama Y, Ellisman MH, Calcutt NA, Cleveland DW (2003) NF-M is an essential target for the myelin-directed "outside-in" signaling cascade that mediates radial axonal growth. In: J Cell Biol, pp 1011-1020.

Gardner EE, Dahl D, Bignami A (1984) Formation of 10-nanometer filaments from the 150K-dalton neurofilament protein in vitro. J Neurosci Res 11:145-155.

143 Gee MA, Heuser JE, Vallee RB (1997) An extended microtubule-binding structure within the dynein motor domain. In: Nature, pp 636-639.

Geisler N, Weber K (1981) Self-assembly in Vitro of the 68,000 molecular weight component of the mammalian neurofilament triplet proteins into intermediate-sized filaments. J Mol Biol 151:565-571.

Gibb BJ, Brion JP, Brownlees J, Anderton BH, Miller CC (1998) Neuropathological abnormalities in transgenic mice harbouring a phosphorylation mutant neurofilament transgene. J Neurochem 70:492- 500.

Glicksman MA, Soppet D, Willard MB (1987) Posttranslational modification of neurofilament polypeptides in rabbit retina. J Neurobiol 18:167-196.

Goldman JE, Yen SH, Chiu FC, Peress NS (1983) Lewy bodies of Parkinson's disease contain neurofilament antigens. Science 221:1082-1084.

Grafstein B, Forman DS (1980) Intracellular transport in neurons. Physiol Rev 60:1167-1283.

Grant P, Pant HC (2000) Neurofilament protein synthesis and phosphorylation. In: J Neurocytol, pp 843-872.

Greene LA (1989) A new neuronal intermediate filament protein. Trends Neurosci 12:228-230.

Gross SP (2004) Hither and yon: a review of bi-directional microtubule-based transport. Phys Biol 1:R1-11.

Gross SP, Welte MA, Block SM, Wieschaus EF (2002) Coordination of opposite- polarity microtubule motors. In: J Cell Biol, pp 715-724.

Harris JW, Moreno S, Shaw G, Mugnaini E (1993) Unusual neurofilament composition in cerebellar unipolar brush neurons. In: J Neurocytol, pp 1039-1059.

He Y, Francis F, Myers KA, Yu W, Black MM, Baas PW (2005) Role of cytoplasmic dynein in the axonal transport of microtubules and neurofilaments. In: J Cell Biol, pp 697-703.

Heidemann SR, Landers JM, Hamborg MA (1981) Polarity orientation of axonal microtubules. In: J Cell Biol, pp 661-665.

Heins S, Wong PC, Müller S, Goldie K, Cleveland DW, Aebi U (1993) The rod domain of NF-L determines neurofilament architecture, whereas the end domains specify filament assembly and network formation. In: J Cell Biol, pp 1517-1533.

144 Helfand BT, Chang L, Goldman RD (2003a) The dynamic and motile properties of intermediate filaments. Annu Rev Cell Dev Biol 19:445-467.

Helfand BT, Loomis P, Yoon M, Goldman RD (2003b) Rapid transport of neural intermediate filament protein. J Cell Sci 116:2345-2359.

Helfand BT, Mendez MG, Pugh J, Delsert C, Goldman RD (2003c) A role for intermediate filaments in determining and maintaining the shape of nerve cells. In: Mol Biol Cell, pp 5069-5081.

Heriot K, Gambetti P, Lasek RJ (1985) Proteins transported in slow components a and b of axonal transport are distributed differently in the transverse plane of the axon. J Cell Biol 100:1167-1172.

Higuchi M, Lee VM, Trojanowski JQ (2002) Tau and axonopathy in neurodegenerative disorders. Neuromolecular Med 2:131-150.

Hirano A, Donnenfeld H, Sasaki S, Nakano I (1984) Fine structural observations of neurofilamentous changes in amyotrophic lateral sclerosis. J Neuropathol Exp Neurol 43:461-470.

Hirokawa N (1982) Cross-linker system between neurofilaments, microtubules, and membranous organelles in frog axons revealed by the quick-freeze, deep-etching method. In: J Cell Biol, pp 129-142.

Hirokawa N (1998a) Kinesin and dynein superfamily proteins and the mechanism of organelle transport. Science 279:519-526.

Hirokawa N (1998b) Kinesin and dynein superfamily proteins and the mechanism of organelle transport. In: Science, pp 519-526.

Hirokawa N, Glicksman MA, Willard MB (1984) Organization of mammalian neurofilament polypeptides within the neuronal cytoskeleton. In: J Cell Biol, pp 1523-1536.

Hirokawa N, Funakoshi ST, Takeda S (1997) Slow axonal transport: the subunit transport model. Trends Cell Biol 7:384-388.

Hirokawa N, Sato-Yoshitake R, Yoshida T, Kawashima T (1990) Brain dynein (MAP1C) localizes on both anterogradely and retrogradely transported membranous organelles in vivo. J Cell Biol 111:1027-1037.

Hirokawa N, Pfister KK, Yorifuji H, Wagner MC, Brady ST, Bloom GS (1989) Submolecular domains of bovine brain kinesin identified by electron microscopy and monoclonal antibody decoration. Cell 56:867-878.

145 Hisanaga S, Hirokawa N (1988) Structure of the peripheral domains of neurofilaments revealed by low angle rotary shadowing. J Mol Biol 202:297-305.

Hisanaga S, Hirokawa N (1989) The effects of dephosphorylation on the structure of the projections of neurofilament. J Neurosci 9:959-966.

Hisanaga S, Hirokawa N (1990) Molecular architecture of the neurofilament. II. Reassembly process of neurofilament L protein in vitro. J Mol Biol 211:871-882.

Hisanaga S, Ikai A, Hirokawa N (1990a) Molecular architecture of the neurofilament. I. Subunit arrangement of neurofilament L protein in the intermediate-sized filament. J Mol Biol 211:857-869.

Hisanaga S, Gonda Y, Inagaki M, Ikai A, Hirokawa N (1990b) Effects of phosphorylation of the neurofilament L protein on filamentous structures. Cell Regul 1:237-248.

Hodge AJ, Adelman WJ, Jr. (1980) The neuroplasmic network in Loligo and Hermissenda neurons. J Ultrastruct Res 70:220-241.

Hoffman PN, Griffin JW, Price DL (1984) Control of axonal caliber by neurofilament transport. In: J Cell Biol, pp 705-714.

Huang JD, Brady ST, Richards BW, Stenolen D, Resau JH, Copeland NG, Jenkins NA (1999) Direct interaction of microtubule- and actin-based transport motors. In: Nature, pp 267-270.

Hurd DD, Saxton WM (1996) Kinesin mutations cause motor neuron disease phenotypes by disrupting fast axonal transport in Drosophila. In: Genetics, pp 1075-1085.

Igisu H, Ohta M, Tabira T, Hosokawa S, Goto I (1975) Giant axonal neuropathy. A clinical entity affecting the central as well as the peripheral nervous system. Neurology 25:717-721.

Jacomy H, Zhu Q, Couillard-Després S, Beaulieu JM, Julien JP (1999) Disruption of type IV intermediate filament network in mice lacking the neurofilament medium and heavy subunits. In: J Neurochem, pp 972-984.

Josephs KA, Holton JL, Rossor MN, Braendgaard H, Ozawa T, Fox NC, Petersen RC, Pearl GS, Ganguly M, Rosa P, Laursen H, Parisi JE, Waldemar G, Quinn NP, Dickson DW, Revesz T (2003) Neurofilament inclusion body disease: a new proteinopathy? Brain 126:2291-2303.

Julien JP, Mushynski WE (1982) Multiple phosphorylation sites in mammalian neurofilament polypeptides. In: J Biol Chem, pp 10467-10470.

146 Julien JP, Mushynski WE (1983) The distribution of phosphorylation sites among identified proteolytic fragments of mammalian neurofilaments. J Biol Chem 258:4019-4025.

Jung C, Shea TB (1999) Regulation of neurofilament axonal transport by phosphorylation in optic axons in situ. Cell Motil Cytoskeleton 42:230-240.

Jung C, Yabe JT, Lee S, Shea TB (2000) Hypophosphorylated neurofilament subunits undergo axonal transport more rapidly than more extensively phosphorylated subunits in situ. In: Cell Motil Cytoskeleton, pp 120-129.

Kanai Y, Dohmae N, Hirokawa N (2004) Kinesin transports RNA: isolation and characterization of an RNA-transporting granule. Neuron 43:513-525.

Kaplan MP, Chin SS, Fliegner KH, Liem RK (1990) Alpha-internexin, a novel neuronal intermediate filament protein, precedes the low molecular weight neurofilament protein (NF-L) in the developing rat brain. J Neurosci 10:2735-2748.

Karki S, Holzbaur EL (1995) Affinity chromatography demonstrates a direct binding between cytoplasmic dynein and the dynactin complex. In: J Biol Chem, pp 28806-28811.

Karki S, LaMonte B, Holzbaur EL (1998) Characterization of the p22 subunit of dynactin reveals the localization of cytoplasmic dynein and dynactin to the midbody of dividing cells. J Cell Biol 142:1023-1034.

King SJ, Schroer TA (2000a) Dynactin increases the processivity of the cytoplasmic dynein motor. Nat Cell Biol 2:20-24.

King SJ, Schroer TA (2000b) Dynactin increases the processivity of the cytoplasmic dynein motor. In: Nat Cell Biol, pp 20-24.

Kong J, Xu ZS (2000) Overexpression of neurofilament subunit NF-L and NF-H extends survival of a mouse model for amyotrophic lateral sclerosis. In: Neurosci Lett, pp 72-74.

Krendel M, Mooseker MS (2005) Myosins: tails (and heads) of functional diversity. In: Physiology (Bethesda, Md), pp 239-251.

Kreplak L, Bar H, Leterrier JF, Herrmann H, Aebi U (2005) Exploring the mechanical behavior of single intermediate filaments. J Mol Biol 354:569- 577.

Kriz J, Zhu Q, Julien JP, Padjen AL (2000a) Electrophysiological properties of axons in mice lacking neurofilament subunit genes: disparity between conduction velocity and axon diameter in absence of NF-H. Brain Res 885:32-44.

147 Kriz J, Meier J, Julien JP, Padjen AL (2000b) Altered ionic conductances in axons of transgenic mouse expressing the human neurofilament heavy gene: A mouse model of amyotrophic lateral sclerosis. Exp Neurol 163:414-421.

Kuczmarski ER, Rosenbaum JL (1979) Studies on the organization and localization of actin and myosin in neurons. In: J Cell Biol, pp 356-371.

Kumar S, Hoh JH (2004) Modulation of repulsive forces between neurofilaments by sidearm phosphorylation. Biochem Biophys Res Commun 324:489- 496.

Kwinter D, Lo KW, Mafi P, Silverman M (2009) Dynactin regulates bidirectional transport of dense-core vesicles in the axon and dendrites of cultured hippocampal neurons. In: Neuroscience, p 10.

Lasek RJ, Garner JA, Brady ST (1984) Axonal transport of the cytoplasmic matrix. In: J Cell Biol, pp 212s-221s.

Lavedan C, Buchholtz S, Nussbaum RL, Albin RL, Polymeropoulos MH (2002) A mutation in the human neurofilament M gene in Parkinson's disease that suggests a role for the cytoskeleton in neuronal degeneration. Neurosci Lett 322:57-61.

Lawrence CJ et al. (2004) A standardized kinesin nomenclature. J Cell Biol 167:19-22.

Lee MK, Cleveland DW (1996) Neuronal intermediate filaments. In: Annu Rev Neurosci, pp 187-217.

Lee MK, Marszalek JR, Cleveland DW (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 MK, Xu ZS, Wong PC, Cleveland DW (1993) Neurofilaments are obligate heteropolymers in vivo. In: J Cell Biol, pp 1337-1350.

Lee VM, Carden MJ, Schlaepfer WW, Trojanowski JQ (1987) Monoclonal antibodies distinguish several differentially phosphorylated states of the two largest rat neurofilament subunits (NF-H and NF-M) and demonstrate their existence in the normal nervous system of adult rats. In: J Neurosci, pp 3474-3488.

Lee VM, Otvos L, Jr., Schmidt ML, Trojanowski JQ (1988a) Alzheimer disease tangles share immunological similarities with multiphosphorylation repeats in the two large neurofilament proteins. Proc Natl Acad Sci U S A 85:7384- 7388.

148 Lee VM, Otvos L, Jr., Carden MJ, Hollosi M, Dietzschold B, Lazzarini RA (1988b) Identification of the major multiphosphorylation site in mammalian neurofilaments. Proc Natl Acad Sci U S A 85:1998-2002.

Lendahl U, Zimmerman LB, McKay RD (1990) CNS stem cells express a new class of intermediate filament protein. Cell 60:585-595.

Leonard DG, Gorham JD, Cole P, Greene LA, Ziff EB (1988) A nerve growth factor-regulated messenger RNA encodes a new intermediate filament protein. J Cell Biol 106:181-193.

Leroy E, Anastasopoulos D, Konitsiotis S, Lavedan C, Polymeropoulos MH (1998) Deletions in the Parkin gene and genetic heterogeneity in a Greek family with early onset Parkinson's disease. Hum Genet 103:424-427.

Leterrier JF, Käs J, Hartwig J, Vegners R, Janmey PA (1996) Mechanical effects of neurofilament cross-bridges. Modulation by phosphorylation, lipids, and interactions with F-actin. In: J Biol Chem, pp 15687-15694.

Letourneau PC (1983) Differences in the organization of actin in the growth cones compared with the neurites of cultured neurons from chick embryos. In: J Cell Biol, pp 963-973.

Lewis SE, Nixon RA (1988) Multiple phosphorylated variants of the high molecular mass subunit of neurofilaments in axons of retinal cell neurons: characterization and evidence for their differential association with stationary and moving neurofilaments. J Cell Biol 107:2689-2701.

Lichtenberg-Kraag B, Mandelkow EM, Biernat J, Steiner B, Schröter C, Gustke N, Meyer HE, Mandelkow E (1992) Phosphorylation-dependent epitopes of neurofilament antibodies on and relationship with Alzheimer tau. In: Proc Natl Acad Sci USA, pp 5384-5388.

Liem RK, Hutchison SB (1982) Purification of individual components of the neurofilament triplet: filament assembly from the 70 000-dalton subunit. Biochemistry 21:3221-3226.

Ligon LA, Steward O (2000) Movement of mitochondria in the axons and dendrites of cultured hippocampal neurons. J Comp Neurol 427:340-350.

Ligon LA, Tokito MK, Finklestein JM, Grossman FE, Holzbaur EL (2004) A direct interaction between cytoplasmic dynein and kinesin I may coordinate motor activity. In: J Biol Chem, pp 19201-19208.

Lim SS, Sammak PJ, Borisy GG (1989) Progressive and spatially differentiated stability of microtubules in developing neuronal cells. In: J Cell Biol, pp 253-263.

149 Lim SS, Edson KJ, Letourneau PC, Borisy GG (1990) A test of microtubule translocation during neurite elongation. In: J Cell Biol, pp 123-130.

Liu Q, Xie F, Siedlak SL, Nunomura A, Honda K, Moreira PI, Zhua X, Smith MA, Perry G (2004) Neurofilament proteins in neurodegenerative diseases. In: Cell Mol Life Sci, pp 3057-3075.

Lorenz T, Willard M (1978) Subcellular fractionation of intra-axonally transport polypeptides in the rabbit visual system. Proc Natl Acad Sci U S A 75:505- 509.

Martin M, Iyadurai SJ, Gassman A, Gindhart JG, Hays TS, Saxton WM (1999) Cytoplasmic dynein, the dynactin complex, and kinesin are interdependent and essential for fast axonal transport. In: Mol Biol Cell, pp 3717-3728.

McKeon FD, Kirschner MW, Caput D (1986) Homologies in both primary and secondary structure between nuclear envelope and intermediate filament proteins. In: Nature, pp 463-468.

Medori R, Autilio-Gambetti L, Monaco S, Gambetti P (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, Autilio-Gambetti L, Jenich H, Gambetti P (1988) Changes in axon size and slow axonal transport are related in experimental diabetic neuropathy. Neurology 38:597-601.

Meier J, Couillard-Despres S, Jacomy H, Gravel C, Julien JP (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.

Melkonian KA, Maier KC, Godfrey JE, Rodgers M, Schroer TA (2007) Mechanism of dynamitin-mediated disruption of dynactin. In: J Biol Chem, pp 19355-19364.

Mercer JA, Seperack PK, Strobel MC, Copeland NG, Jenkins NA (1991) Novel myosin heavy chain encoded by murine dilute coat colour locus. In: Nature, pp 709-713.

Mersiyanova IV, Perepelov AV, Polyakov AV, Sitnikov VF, Dadali EL, Oparin RB, Petrin AN, Evgrafov OV (2000) A new variant of Charcot-Marie-Tooth disease type 2 is probably the result of a mutation in the neurofilament- light gene. Am J Hum Genet 67:37-46.

Metuzals J, Tasaki I (1978) Subaxolemmal filamentous network in the giant nerve fiber of the squid (Loligo pealei L.) and its possible role in excitability. In: J Cell Biol, pp 597-621.

150 Miki H, Setou M, Kaneshiro K, Hirokawa N (2001) All kinesin superfamily protein, KIF, genes in mouse and human. Proc Natl Acad Sci U S A 98:7004-7011.

Monaco S, Autilio-Gambetti L, Zabel D, Gambetti P (1985) Giant axonal neuropathy: acceleration of neurofilament transport in optic axons. Proc Natl Acad Sci U S A 82:920-924.

Monteiro MJ, Hoffman PN, Gearhart JD, Cleveland DW (1990) Expression of NF- L in both neuronal and nonneuronal cells of transgenic mice: increased neurofilament density in axons without affecting caliber. In: J Cell Biol, pp 1543-1557.

Morris JR, Lasek RJ (1982) Stable polymers of the axonal cytoskeleton: the axoplasmic ghost. In: J Cell Biol, pp 192-198.

Morris JR, Lasek RJ (1984) Monomer-polymer equilibria in the axon: direct measurement of tubulin and actin as polymer and monomer in axoplasm. In: J Cell Biol, pp 2064-2076.

Morris RL, Hollenbeck PJ (1993) The regulation of bidirectional mitochondrial transport is coordinated with axonal outgrowth. In: J Cell Sci, pp 917-927.

Morris RL, Hollenbeck PJ (1995) Axonal transport of mitochondria along microtubules and F-actin in living vertebrate neurons. In: J Cell Biol, pp 1315-1326.

Mukai H, Toshimori M, Shibata H, Kitagawa M, Shimakawa M, Miyahara M, Sunakawa H, Ono Y (1996) PKN associates and phosphorylates the head-rod domain of neurofilament protein. J Biol Chem 271:9816-9822.

Müller MJ, Klumpp S, Lipowsky R (2008) Tug-of-war as a cooperative mechanism for bidirectional cargo transport by molecular motors. In: Proc Natl Acad Sci USA, pp 4609-4614.

Munoz DG, Greene C, Perl DP, Selkoe DJ (1988) Accumulation of phosphorylated neurofilaments in anterior horn motoneurons of amyotrophic lateral sclerosis patients. J Neuropathol Exp Neurol 47:9-18.

Murayama S, Suzuki I, Nagase M, Shingaki S, Kawasaki T, Nakajima T, Fukushima M, Ishiki T (1988) Chondrosarcoma of the mandible. Report of case and a survey of 23 cases in the Japanese literature. J Craniomaxillofac Surg 16:287-292.

Murray JW, Bananis E, Wolkoff AW (2000) Reconstitution of ATP-dependent movement of endocytic vesicles along microtubules in vitro: an oscillatory bidirectional process. Mol Biol Cell 11:419-433.

151 Myers MW, Lazzarini RA, Lee VM, Schlaepfer WW, Nelson DL (1987) The human mid-size neurofilament subunit: a repeated protein sequence and the relationship of its gene to the intermediate filament gene family. In: EMBO J, pp 1617-1626.

Napolitano EW, Chin SS, Colman DR, Liem RK (1987) Complete amino acid sequence and in vitro expression of rat NF-M, the middle molecular weight neurofilament protein. In: J Neurosci, pp 2590-2599.

Navone F, Niclas J, Hom-Booher N, Sparks L, Bernstein HD, McCaffrey G, Vale RD (1992) Cloning and expression of a human kinesin heavy chain gene: interaction of the COOH-terminal domain with cytoplasmic microtubules in transfected CV-1 cells. In: J Cell Biol, pp 1263-1275.

Nguyen MD, Larivière RC, Julien JP (2000) Reduction of axonal caliber does not alleviate motor neuron disease caused by mutant superoxide dismutase 1. In: Proc Natl Acad Sci USA, pp 12306-12311.

Nguyen MD, Shu T, Sanada K, Lariviere RC, Tseng HC, Park SK, Julien JP, Tsai LH (2004) A NUDEL-dependent mechanism of neurofilament assembly regulates the integrity of CNS neurons. Nat Cell Biol 6:595-608.

Nixon RA, Paskevich PA, Sihag RK, Thayer CY (1994) Phosphorylation on carboxyl terminus domains of neurofilament proteins in retinal ganglion cell neurons in vivo: influences on regional neurofilament accumulation, interneurofilament spacing, and axon caliber. In: J Cell Biol, pp 1031- 1046.

Norris EH, Giasson BI, Lee VM (2004) Alpha-synuclein: normal function and role in neurodegenerative diseases. Curr Top Dev Biol 60:17-54.

O'Connell CB, Tyska MJ, Mooseker MS (2007) Myosin at work: motor adaptations for a variety of cellular functions. In: Biochim Biophys Acta, pp 615-630.

Ogura T, Whiteheart SW, Wilkinson AJ (2004) Conserved arginine residues implicated in ATP hydrolysis, nucleotide-sensing, and inter-subunit interactions in AAA and AAA+ ATPases. J Struct Biol 146:106-112.

Ohara O, Gahara Y, Miyake T, Teraoka H, Kitamura T (1993) Neurofilament deficiency in quail caused by nonsense mutation in neurofilament-L gene. In: J Cell Biol, pp 387-395.

Okabe S, Hirokawa N (1992) Differential behavior of photoactivated microtubules in growing axons of mouse and frog neurons. In: J Cell Biol, pp 105-120.

152 Pappolla MA (1986) Lewy bodies of Parkinson's disease. Immune electron microscopic demonstration of neurofilament antigens in constituent filaments. Arch Pathol Lab Med 110:1160-1163.

Parysek LM, Chisholm RL, Ley CA, Goldman RD (1988) A type III intermediate filament gene is expressed in mature neurons. Neuron 1:395-401.

Paschal BM, Vallee RB (1987) Retrograde transport by the microtubule- associated protein MAP 1C. Nature 330:181-183.

Patterson GH, Lippincott-Schwartz J (2002) A photoactivatable GFP for selective photolabeling of proteins and cells. Science 297:1873-1877.

Perez-Olle R, Leung CL, Liem RK (2002) Effects of Charcot-Marie-Tooth-linked mutations of the neurofilament light subunit on intermediate filament formation. J Cell Sci 115:4937-4946.

Perrot R, Lonchampt P, Peterson AC, Eyer J (2007) Axonal neurofilaments control multiple fiber properties but do not influence structure or spacing of nodes of Ranvier. J Neurosci 27:9573-9584.

Perrot R, Berges R, Bocquet A, Eyer J (2008) Review of the multiple aspects of neurofilament functions, and their possible contribution to neurodegeneration. In: Mol Neurobiol, pp 27-65.

Pfister KK, Salata MW, Dillman JF, Torre E, Lye RJ (1996a) Identification and developmental regulation of a neuron-specific subunit of cytoplasmic dynein. In: Mol Biol Cell, pp 331-343.

Pfister KK, Salata MW, Dillman JF, 3rd, Vaughan KT, Vallee RB, Torre E, Lye RJ (1996b) Differential expression and phosphorylation of the 74-kDa intermediate chains of cytoplasmic dynein in cultured neurons and glia. J Biol Chem 271:1687-1694.

Portier MM, de Nechaud B, Gros F (1983) Peripherin, a new member of the intermediate filament protein family. Dev Neurosci 6:335-344.

Quintyne NJ, Schroer TA (2002) Distinct cell cycle-dependent roles for dynactin and dynein at centrosomes. In: J Cell Biol, pp 245-254.

Quintyne NJ, Gill SR, Eckley DM, Crego CL, Compton DA, Schroer TA (1999) Dynactin is required for microtubule anchoring at centrosomes. In: J Cell Biol, pp 321-334.

Rammensee S, Janmey PA, Bausch AR (2007) Mechanical and structural properties of in vitro neurofilament hydrogels. In: Eur Biophys J, pp 661- 668.

153 Rao MV, Houseweart MK, Williamson TL, Crawford TO, Folmer J, Cleveland DW (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. In: J Cell Biol, pp 171-181.

Richards TA, Cavalier-Smith T (2005) Myosin domain evolution and the primary divergence of eukaryotes. In: Nature, pp 1113-1118.

Rodriguez OC, Cheney RE (2002) Human myosin-Vc is a novel class V myosin expressed in epithelial cells. In: J Cell Sci, pp 991-1004.

Rogers SL, Tint IS, Fanapour PC, Gelfand VI (1997) Regulated bidirectional motility of melanophore pigment granules along microtubules in vitro. Proc Natl Acad Sci U S A 94:3720-3725.

Ross J, Olmsted JB, Rosenbaum JL (1975) The ultrastructure of mouse neuroblastoma cells in tissue culture. Tissue Cell 7:107-135.

Ross JL, Wallace KE, Shuman H, Goldman YE, Holzbaur EL (2006) Processive bidirectional motion of dynein-dynactin complexes in vitro. In: Nat Cell Biol, pp 562-570.

Roy S, Zhang B, Lee VM, Trojanowski JQ (2005) Axonal transport defects: a common theme in neurodegenerative diseases. In: Acta Neuropathol, pp 5-13.

Roy S, Coffee P, Smith G, Liem RK, Brady ST, Black MM (2000) Neurofilaments are transported rapidly but intermittently in axons: implications for slow axonal transport. J Neurosci 20:6849-6861.

Sabry J, O'Connor TP, Kirschner MW (1995) Axonal transport of tubulin in Ti1 pioneer neurons in situ. In: Neuron, pp 1247-1256.

Sakaguchi T, Okada M, Kitamura T, Kawasaki K (1993) Reduced diameter and conduction velocity of myelinated fibers in the sciatic nerve of a neurofilament-deficient mutant quail. Neurosci Lett 153:65-68.

Sanchez I, Hassinger L, Sihag RK, Cleveland DW, Mohan P, Nixon RA (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.

Sánchez I, Hassinger L, Paskevich PA, Shine HD, Nixon RA (1996) Oligodendroglia regulate the regional expansion of axon caliber and local accumulation of neurofilaments during development independently of myelin formation. In: J Neurosci, pp 5095-5105.

154 Saxton WM (1994) Isolation and analysis of microtubule motor proteins. Methods Cell Biol 44:279-288.

Schafer DA, Gill SR, Cooper JA, Heuser JE, Schroer TA (1994) Ultrastructural analysis of the dynactin complex: an actin-related protein is a component of a filament that resembles F-actin. J Cell Biol 126:403-412.

Schmidt ML, Lee VM, Trojanowski JQ (1990) Relative abundance of tau and neurofilament epitopes in hippocampal neurofibrillary tangles. In: Am J Pathol, pp 1069-1075.

Schmidt RE, Plurad SB (1985) Ultrastructural appearance of intentionally frustrated axonal regeneration in rat sciatic nerve. J Neuropathol Exp Neurol 44:130-146.

Schmidt RE, Beaudet LN, Plurad SB, Dorsey DA (1997) Axonal cytoskeletal pathology in aged and diabetic human sympathetic autonomic ganglia. Brain Res 769:375-383.

Schroer TA, Bingham JB, Gill SR (1996) Actin-related protein 1 and cytoplasmic dynein-based motility - what's the connection? Trends Cell Biol 6:212-215.

Schwob JE, Farber NB, Gottlieb DI (1986) Neurons of the olfactory epithelium in adult rats contain vimentin. J Neurosci 6:208-217.

Scott D, Smith KE, O'Brien BJ, Angelides KJ (1985) Characterization of mammalian neurofilament triplet proteins. Subunit stoichiometry and morphology of native and reconstituted filaments. J Biol Chem 260:10736- 10747.

Scott JN, Clark AW, Zochodne DW (1999) Neurofilament and tubulin in progressive experimental diabetes: failure of synthesis and export by sensory neurons. Brain 122 ( Pt 11):2109-2118.

Selkoe DJ, Ihara Y, Salazar FJ (1982) Alzheimer's disease: insolubility of partially purified paired helical filaments in sodium dodecyl sulfate and urea. Science 215:1243-1245.

Shah JV, Cleveland DW (2002) Slow axonal transport: fast motors in the slow lane. In: Curr Opin Cell Biol, pp 58-62.

Shah JV, Flanagan LA, Janmey PA, Leterrier JF (2000) Bidirectional translocation of neurofilaments along microtubules mediated in part by dynein/dynactin. In: Mol Biol Cell, pp 3495-3508.

Shaner NC, Campbell RE, Steinbach PA, Giepmans BN, Palmer AE, Tsien RY (2004) Improved monomeric red, orange and yellow fluorescent proteins

155 derived from Discosoma sp. red fluorescent protein. Nat Biotechnol 22:1567-1572.

Shaw G, Weber K (1983) The structure and development of the rat retina: an immunofluorescence microscopical study using antibodies specific for intermediate filament proteins. Eur J Cell Biol 30:219-232.

Shaw G, Weber K (1984) The intermediate filament complement of the retina: a comparison between different mammalian species. Eur J Cell Biol 33:95- 104.

Shea TB, Yabe JT, Ortiz D, Pimenta A, Loomis P, Goldman RD, Amin N, Pant HC (2004) Cdk5 regulates axonal transport and phosphorylation of neurofilaments in cultured neurons. J Cell Sci 117:933-941.

Shetty KT, Link WT, Pant HC (1993) cdc2-like kinase from rat spinal cord specifically phosphorylates KSPXK motifs in neurofilament proteins: isolation and characterization. In: Proc Natl Acad Sci USA, pp 6844-6848.

Shubeita G, Tran S, Xu J, Vershinin M, Cermelli S, Cotton S, Welte MA, Gross SP (2008) Consequences of motor copy number on the intracellular transport of kinesin-1-driven lipid droplets. In: Cell, pp 1098-1107.

Sihag RK, Nixon RA (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. In: J Biol Chem, pp 18861-18867.

Sihag RK, Jaffe H, Nixon RA, Rong X (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. In: J Neurochem, pp 491-499.

Skre H (1978) Current research in neuro-epidemiology. Some main trends. Acta Neurol Scand Suppl 67:11-36.

Spillantini MG, Crowther RA, Jakes R, Hasegawa M, Goedert M (1998) alpha- Synuclein in filamentous inclusions of Lewy bodies from Parkinson's disease and dementia with lewy bodies. In: Proc Natl Acad Sci USA, pp 6469-6473.

Spooner BS, Holladay CR (1981) Distribution of tubulin and actin in neurites and growth cones of differentiating nerve cells. In: Cell Motil, pp 167-178.

Spooner BS, Holladay CR, Bright GR (1982) Immunofluorescence comparisons of anti-actin specificity. In: Eur J Cell Biol, pp 115-121.

156 Steffen W, Hodgkinson JL, Wiche G (1996) Immunogold localisation of the intermediate chain within the protein complex of cytoplasmic dynein. In: J Struct Biol, pp 227-235.

Steinert PM, Roop DR (1988) Molecular and cellular biology of intermediate filaments. In: Annu Rev Biochem, pp 593-625.

Sternberger NH, Sternberger LA, Ulrich J (1985) Aberrant neurofilament phosphorylation in Alzheimer disease. Proc Natl Acad Sci U S A 82:4274- 4276.

Strobel MC, Seperack PK, Copeland NG, Jenkins NA (1990) Molecular analysis of two mouse dilute locus deletion mutations: spontaneous dilute lethal20J and radiation-induced dilute prenatal lethal Aa2 alleles. In: Mol Cell Biol, pp 501-509.

Suomalainen M, Nakano MY, Keller S, Boucke K, Stidwill RP, Greber UF (1999) Microtubule-dependent plus- and minus end-directed motilities are competing processes for nuclear targeting of adenovirus. In: J Cell Biol, pp 657-672.

Takeda S, Funakoshi T, Hirokawa N (1995) Tubulin dynamics in neuronal axons of living zebrafish embryos. In: Neuron, pp 1257-1264.

Terada S (2003) Where does slow axonal transport go? Neurosci Res 47:367- 372.

Terada S, Hirokawa N (2000) Moving on to the cargo problem of microtubule- dependent motors in neurons. Curr Opin Neurobiol 10:566-573.

Toyoshima I, Komiya Y (1995) Phosphorylation and transport of neurofilament proteins in the rat spinal ganglion. Neurosci Lett 189:69-72.

Trojanowski JQ, Mattson MP (2003) Overview of protein aggregation in single, double, and triple neurodegenerative brain amyloidoses. Neuromolecular Med 4:1-6.

Troy CM, Muma NA, Greene LA, Price DL, Shelanski ML (1990) Regulation of peripherin and neurofilament expression in regenerating rat motor neurons. Brain Res 529:232-238.

Tu PH, Raju P, Robinson KA, Gurney ME, Trojanowski JQ, Lee VM (1996) Transgenic mice carrying a human mutant superoxide dismutase transgene develop neuronal cytoskeletal pathology resembling human amyotrophic lateral sclerosis lesions. Proc Natl Acad Sci U S A 93:3155- 3160.

157 Tytell M, Black MM, Garner JA, Lasek RJ (1981) Axonal transport: each major rate component reflects the movement of distinct macromolecular complexes. In: Science, pp 179-181.

Uchida A, Brown A (2004) Arrival, reversal, and departure of neurofilaments at the tips of growing axons. In: Mol Biol Cell, pp 4215-4225.

Uchida A, Alami NH, Brown A (2009) Tight Functional Coupling of Kinesin-1A and Dynein Motors in the Bidirectional Transport of Neurofilaments. Mol Biol Cell.

Uchikado H, Shaw G, Wang DS, Dickson DW (2005) Screening for neurofilament inclusion disease using alpha-internexin immunohistochemistry. Neurology 64:1658-1659.

Vale RD, Reese TS, Sheetz MP (1985) Identification of a novel force-generating protein, kinesin, involved in microtubule-based motility. Cell 42:39-50.

Vaughan KT, Vallee RB (1995) Cytoplasmic dynein binds dynactin through a direct interaction between the intermediate chains and p150Glued. In: J Cell Biol, pp 1507-1516.

Vaughan PS, Leszyk JD, Vaughan KT (2001) Cytoplasmic dynein intermediate chain phosphorylation regulates binding to dynactin. In: J Biol Chem, pp 26171-26179.

Veeranna, Lee JH, Pareek TK, Jaffee H, Boland B, Vinod KY, Amin N, Kulkarni AB, Pant HC, Nixon RA (2008) Neurofilament tail phosphorylation: identity of the RT-97 phosphoepitope and regulation in neurons by cross-talk among proline-directed kinases. In: J Neurochem, pp 35-49.

Verbecke G, Molenberghs G (2000) Linear mixexd models for longitudinal data: New York: Springer.

Vogel P, Gabriel M, Goebel HH, Dyck PJ (1985) Hereditary motor sensory neuropathy type II with neurofilament accumulation: new finding or new disorder? Ann Neurol 17:455-461.

Wagner OI, Ascaño J, Tokito MK, Leterrier JF, Janmey PA, Holzbaur EL (2004) The interaction of neurofilaments with the microtubule motor cytoplasmic dynein. In: Mol Biol Cell, pp 5092-5100.

Wang L, Brown A (2001) Rapid intermittent movement of axonal neurofilaments observed by fluorescence photobleaching. In: Mol Biol Cell, pp 3257-3267.

Wang L, Ho CL, Sun D, Liem RK, Brown A (2000) Rapid movement of axonal neurofilaments interrupted by prolonged pauses. In: Nat Cell Biol, pp 137- 141.

158 Waterman-Storer CM, Karki S, Holzbaur EL (1995) The p150Glued component of the dynactin complex binds to both microtubules and the actin-related protein centractin (Arp-1). Proc Natl Acad Sci U S A 92:1634-1638.

Waterman-Storer CM, Karki SB, Kuznetsov SA, Tabb JS, Weiss DG, Langford GM, Holzbaur EL (1997) The interaction between cytoplasmic dynein and dynactin is required for fast axonal transport. In: Proc Natl Acad Sci USA, pp 12180-12185.

Watson DF, Hoffman PN, Fittro KP, Griffin JW (1989) Neurofilament and tubulin transport slows along the course of mature motor axons. Brain Res 477:225-232.

Weiss P, Hiscoe HB (1948) Experiments on the mechanism of nerve growth. J Exp Zool 107:315-395.

Welte MA, Gross SP, Postner M, Block SM, Wieschaus EF (1998) Developmental regulation of vesicle transport in Drosophila embryos: forces and kinetics. In: Cell, pp 547-557.

Welte MA, Cermelli S, Griner J, Viera A, Guo Y, Kim DH, Gindhart JG, Gross SP (2005) Regulation of lipid-droplet transport by the perilipin homolog LSD2. Curr Biol 15:1266-1275.

Williamson TL, Cleveland DW (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.

Williamson TL, Bruijn LI, Zhu Q, Anderson KL, Anderson SD, Julien JP, Cleveland DW (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. In: Proc Natl Acad Sci USA, pp 9631-9636.

Wong PC, Marszalek JR, Crawford TO, Xu ZS, Hsieh ST, Griffin JW, Cleveland DW (1995) Increasing neurofilament subunit NF-M expression reduces axonal NF-H, inhibits radial growth, and results in neurofilamentous accumulation in motor neurons. In: J Cell Biol, pp 1413-1422.

Wu C, Ramirez A, Cui B, Ding J, Delcroix JD, Valletta JS, Liu JJ, Yang Y, Chu S, Mobley WC (2007) A functional dynein-microtubule network is required for NGF signaling through the Rap1/MAPK pathway. Traffic 8:1503-1520.

Wu X, Bowers B, Rao K, Wei Q, Hammer JA, 3rd (1998) Visualization of melanosome dynamics within wild-type and dilute melanocytes suggests a paradigm for myosin V function In vivo. J Cell Biol 143:1899-1918.

159 Xia C, Rahman A, Yang Z, Goldstein LS (1998) Chromosomal localization reveals three kinesin heavy chain genes in mouse. Genomics 52:209-213.

Xia CH, Roberts EA, Her LS, Liu X, Williams DS, Cleveland DW, Goldstein LS (2003) Abnormal neurofilament transport caused by targeted disruption of neuronal kinesin heavy chain KIF5A. In: J Cell Biol, pp 55-66.

Xu Z, Cork LC, Griffin JW, Cleveland DW (1993) Increased expression of neurofilament subunit NF-L produces morphological alterations that resemble the pathology of human motor neuron disease. In: Cell, pp 23- 33.

Xu ZS, Liu WS, Willard MB (1992) Identification of six phosphorylation sites in the COOH-terminal tail region of the rat neurofilament protein M. J Biol Chem 267:4467-4471.

Xu ZS, Marszalek JR, Lee MK, Wong PC, Folmer J, Crawford TO, Hsieh ST, Griffin JW, Cleveland DW (1996) Subunit composition of neurofilaments specifies axonal diameter. In: J Cell Biol, pp 1061-1069.

Yabe JT, Pimenta A, Shea TB (1999) Kinesin-mediated transport of neurofilament protein oligomers in growing axons. In: J Cell Sci, pp 3799- 3814.

Yabe JT, Jung C, Chan WK, Shea TB (2000) Phospho-dependent association of neurofilament proteins with kinesin in situ. Cell Motil Cytoskeleton 45:249- 262.

Yabe JT, Chan WK, Chylinski TM, Lee S, Pimenta AF, Shea TB (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 JT, Chylinski T, Wang FS, Pimenta A, Kattar SD, Linsley MD, Chan WK, Shea TB (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.

Yagihashi S, Kamijo M, Watanabe K (1990) Reduced myelinated fiber size correlates with loss of axonal neurofilaments in peripheral nerve of chronically streptozotocin diabetic rats. Am J Pathol 136:1365-1373.

Yamasaki H, Itakura C, Mizutani M (1991) Hereditary hypotrophic axonopathy with neurofilament deficiency in a mutant strain of the Japanese quail. Acta Neuropathol 82:427-434.

160 Yamasaki H, Bennett GS, Itakura C, Mizutani M (1992) Defective expression of neurofilament protein subunits in hereditary hypotrophic axonopathy of quail. Lab Invest 66:734-743.

Yan Y, Brown A (2005) Neurofilament polymer transport in axons. In: J Neurosci, pp 7014-7021.

Yan Y, Jensen K, Brown A (2007) The polypeptide composition of moving and stationary neurofilaments in cultured sympathetic neurons. In: Cell Motil Cytoskeleton, pp 299-309.

Yang Y, Allen E, Ding J, Wang W (2007) Giant axonal neuropathy. Cell Mol Life Sci 64:601-609.

Yuan A, Rao MV, Sasaki T, Chen Y, Kumar A, Veeranna, Liem RK, Eyer J, Peterson AC, Julien JP, Nixon RA (2006) Alpha-internexin is structurally and functionally associated with the neurofilament triplet proteins in the mature CNS. In: J Neurosci, pp 10006-10019.

Zhang B, Tu P, Abtahian F, Trojanowski JQ, Lee VM (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.

Zhao LP, Koslovsky JS, Reinhard J, Bähler M, Witt AE, Provance DW, Mercer JA (1996) Cloning and characterization of myr 6, an unconventional myosin of the dilute/myosin-V family. In: Proc Natl Acad Sci USA, pp 10826-10831.

Zhu Q, Couillard-Despres S, Julien JP (1997) Delayed maturation of regenerating myelinated axons in mice lacking neurofilaments. Exp Neurol 148:299-316.

Zhu Q, Lindenbaum M, Levavasseur F, Jacomy H, Julien JP (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. In: J Cell Biol, pp 183-193.

Zochodne DW, Sun HS, Cheng C, Eyer J (2004a) Accelerated diabetic neuropathy in axons without neurofilaments. In: Brain, pp 2193-2200.

Zochodne DW, Sun HS, Cheng C, Eyer J (2004b) Accelerated diabetic neuropathy in axons without neurofilaments. Brain 127:2193-2200.

161