University of Connecticut OpenCommons@UConn
Doctoral Dissertations University of Connecticut Graduate School
10-24-2013 Localization and Dynamics of Myosin Vb in Mammalian Cells Eun-Ji Kim [email protected]
Follow this and additional works at: https://opencommons.uconn.edu/dissertations
Recommended Citation Kim, Eun-Ji, "Localization and Dynamics of Myosin Vb in Mammalian Cells" (2013). Doctoral Dissertations. 250. https://opencommons.uconn.edu/dissertations/250
Localization and Dynamics of Myosin Vb in Mammalian Cells
Eun-Ji Kim, Ph.D.
University of Connecticut, 2013
Abstract
The myosin superfamily is a large and diverse family of proteins that are
capable of binding to actin filaments and generating mechanical forces by
hydrolyzing ATP. Myosins perform a diverse set of functions in cell, including
endocytosis, exocytosis, movement of pigment granules or mRNA, and cell
motility1. More specifically, Class-V myosins are a group of barbed-end directed
motors that have been implicated in organelle transport. For example, in vitro
motility assays suggest that myosin Va is a processive motor with a step size of
36 nm2-4 and moves at the speed of 0.6 µm/sec, which is consistent with a
functional role in actin-based transport process. Although the properties of myosin
V in vitro are relatively well-studied, much less is understood about the motility
and protein-protein interaction of myosin V in vivo.
Previous studies of molecular motor proteins in vivo have analyzed
intracellular cargoes, which are decorated with multiple types of motor proteins5,6,
making it difficult to deduce the exact dynamics and the function of a specific
motor protein. In this work, we carried out experiments to analyze the localization
and the dynamics of myosin Vb in live mammalian cells at both the ensemble
level and the level of individual myosin Vb molecules. We found that: 1) motor
molecules localize and accumulate at regions enriched in F-actin
i Eun-Ji Kim-University of Connecticut, 2013
barbed-ends; 2) individual motor molecules have a high stalling rate when bound
to F-actin and do not undergo continuous long-distance movement in cells and 3)
over-expression of myosin Vb motor domains promotes filopodia growth. Based
on these observations, we propose a novel cellular function of myosin Vb
proteins: myosin Vb facilitates linear F-actin growth by moving towards, and
eventually binding to, the barbed-ends of individual actin filaments, which in turn
prevents capping of the F-actin. Therefore, our study provides new insights into
the mechanism of myosin’s function in filopodia formation.
ii
Localization and Dynamics of Myosin Vb in Mammalian Cells
Eun-Ji Kim, Ph.D.
B.S. Chonnam National University, 2004
M.A. Central Connecticut State University, 2008
A Dissertation
Submitted in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
at the
University of Connecticut
[2013]
iii
Copyright by
Eun-Ji Kim
[2013]
iv APPROVAL PAGE
Doctor of Philosophy Dissertation
Localization and Dynamics of Myosin Vb in Mammalian Cells.
Presented by
Eun-Ji Kim, B.S.,M.A.
Major Advisor ______
Dr. Ji Yu
Associate Advisor ______
Dr. John H. Carson
Associate Advisor ______
Dr. Vladimir Rodionov
Associate Advisor ______
Dr. Srdjan D. Antic
University of Connecticut
[2013]
v Acknowledgements
Five years have passed since I joined Dr. Ji Yu’s lab. While in Dr. Ji Yu lab, I met a lot of amazing people including Dr. Sulagna Das, Qingfen Yang, Dr. Jingqiao Zhang, Dr.
Veda Tatarvarty, Dr. DongMyung Oh, Erika Hoyos Ramirez, Olena Marchenko, Cliff
Locke, Ahmed Elmokadem, Yei feng. They were extremely helpful and patient with me.
Without their help and support, my life during PhD training would have been more difficult. I would like to thank Dr. Ji Yu, I feel very lucky to be under his mentorship. He is a brilliant scientist with a wide range of knowledge including Chemistry, Biology,
Physics, Mathematics, Programming, etc. His brain blowing ideas towards solving problems in biology kept me awake overnight with curiosity.
I would also like to thank my committee members, who were very tough, but very cool. I am a very lucky person to have my committee member, Dr. Vladimir Rodionov whose passion for science has directly transformed me during my PhD training. He has increased my excitement in and helped me greatly in understanding the functioning of motor proteins due to his expertise and experience. He invited amazing scientists whose works are directly related to my work, I was able to meet them and learn from them. I feel very lucky to have him as my committee member. I would like to thank Dr.
John Carson, who is very supportive in the opinions of the students and provide continuous motivations while increasing our curiosity by asking a lot of tough questions.
His attitude towards science and mentorship had a huge impact on me personally. I feel very lucky to have an opportunity to get close to him. I would like to thank Dr. Srdjan
Antic, his one on one approach to students, giving priceless feedback and sincere
vi empathy about our performance deemed very useful, and I am very lucky to have met him.
I would like to thank Sofya Borinskaya, as her research talks increased my thesis progress enormously. She is a great peer and we working on similar topic helped to increased my scientific knowledge. I would like to thank Dr. Akeisha Belgrave, I start
PhD training with her, we went through a lot of things together, and are finally finishing up at the same time. Cheers to both of us. I would like to thank Dr. Yi Wu and Dr. Taofei
Yin. They helped me with lot of things including reagents and very valuable feedback.
I would like to thank Jian-Ping Haung and Dr. Xin-Ming Ma who give me a lot of help and mentorship. Special thanks to JP!
I would like to thank the other people in CCAM including Dr. Raquell Homes, Dr. Jing
Yang, Dr. Cibelle Falkenberg. Dr. Ann Cowan, Dr. Leslie Loew, Dr. Boris Slepchenko,
Dr. Igor Novak, Dr. Michael Blinov, Dr. Ion Moraru, Dr. Charles Wogemuth, Dutton Jeff, and CCAM students Rene Norman, Abhijit Deb Roy.
Lastly, I would like to thank my parents, and my brother. My parents are my biggest support, my role models, my best friends. I really appreciate their supports and endless love. I would like to thank my husband (my unofficial committee member), who taught me a lot of things and helped me a lot to finish up. He, dedicated his time to me, helped my unstable mood, tolerated my busy life and supported my dream.
vii TABLE OF CONTENTS
ACKNOWLEDGEMENTS………………………………………………………………… vi
LIST OF FIGURES ……………………………………………………………………….. xi
ABBREVIATIONS…………………………………………………………………………. xii
Chapter 1..……………………………………………………………………………… 1
1.1 Introduction to Myosin Motor Proteins………...………………………….…….. 1
1.1.1 Structure and function of Myosins………………...…………………………… 1
1.1.2 In vitro properties of Myosin Vb …………………..……………………………. 3
1.1.3 In vivo function of Myosin Vb…………………..…………………………...... 5
1.2 Myosin Vb’s role in intracellular transport………………..…………………….. 7
1.2.1 Processivity and in vivo cargo transport……………………..………………… 7
1.2.2 Regulation of Myosin Vb cargo binding………………………………..………. 8
1.3 Myosin Vb’s role in filopodia formation ………………………………….……… 9
1.3.1 Actin cytoskeleton ……………………………………………...………………... 9
1.3.2 Mechanism of filopodia formation………………………………………..……... 11
1.3.3 Myosin Vb is required for microvilli formation……………………….………… 13
1.3.4 Other myosins implicated in filopodia formation.……………………………… 14
1.4 Single molecule imaging of Single Myosin Motor in cells………………...….. 15
1.5 Thesis hypothesis and objectives……………………………………………..….. 16
Chapter 2……………………………………………………….….. 18
2.1 Abstract………………………………………………………………………………… 18
viii 2.2 Introduction…………………………………………………………………………… 18
2.3 Material and method…………………………………………………………………. 20
2.3.1 Cloning and Mutagenesis…………………………………………………….….. 20
2.3.2 Cell culture and transfection……………………………………………………… 21
2.3.3 Drug treatment…………………………………………………………………….. 21
2.3.4 Western blotting…………………………………………………………………… 21
2.3.5 Single molecule Imaging and analysis………………………………………….. 22
2.3.6 MSD analysis………………………………………………………………………. 23
2.4 Result…………………………………………………………………………………… 23
2.4.1 Myosin Vb localizes to cell edge………………………………………………… 23
2.4.2 Cell edge localization of myosin Vb depends on motor activity……………… 25
2.4.3 Single molecule quantification of myosin Vb motility in live cells…………….. 25
2.4.4 Myosin Vb processivity…………………………………………………………… 28
2.4.5 Rif expression increase dissociation rate of myosin molecules from cortical
actin………………………………………………………………………………… 29
2.4.6 Myosin Vb tethering at actin barbed-end……………………………………….. 30
2.5 Discussion…………………………………………………………………………….. 32
Chapter 3……………………………………………………….….. 34
3.1 Abstract………………………………………………………………………………… 34
3.2 Introduction…………………………………………………………………………… 34
3.3 Material and methods……………………………………………………………….. 36
3.3.1 Cloning and constructs…………………………………………………………… 36
3.3.2 Cell culture and transfection……………………………………………………… 37
ix 3.3.3 Immunofluorescence and antibodies……………………………………………. 37
3.3.4 Total internal reflective fluorescent (TIRF) imaging and live cell imaging…... 38
3.3.5 Confocal imaging and fluorescent recovery after photobleaching (FRAP)…. 38
3.4 Results………………………………………………………………………………..... 39
3.4.1 Myosin Vb induce filopodia formation………………………………………...... 39
3.4.2 Myosin Vb forms tight focus at the growing filopodia tip………………..…...... 40
3.4.3 SptPALM single molecule tracking of myosin Vb within filopodia……….…… 41
3.4.4 Myosin Vb competes with other barbed-end binding proteins for actin
binding………………………………………………………………………….…... 42
3.5 Discussion…………………………………………………………………………...... 44
Chapter 4…………………………………………………………... 47
4.1 Summary of findings…………………………………………………………….…... 47
4.1.1 Individual behavior of Myosin Vb Motors in cell is tethered at the barbed
end of the filaments……………………………………………………………….. 47
4.1.2 Myosin Vb motors rearrange the actin cytoskeleton organization………...... 48
4.2 Future directions………………………………………………………………...…… 48
4.2.1 Single molecule assay of Myosin V paralogues: Myosin Va and Vc……...... 48
4.2.2 Mechanism that Myosin Vb induce filopodia……………………………...... 49
Figures……………………………………………………………………………………… 50
Tables………………………………………………………………………………….……. 70
References………………………………………………………………………….……… 71
x LIST OF FIGURES
Figure 1-1. Models for myosin V processivity……………………………………..………….. 50
Figure 1-2. Tug of war mechanism………….…………………………………….…………… 51
Figure 2-1. PALM microscope set up…………………………………………….……………. 52
Figure 2-2. Normalized excitation and emission spectra of WT Eos FP………..…………. 53
Figure 2-3. Expected mode of motion of myosin V……………………………..……………. 54
Figure 2-4. Western blot of myosin V……………………..………………………...…………. 55
Figure 2-5. Myosin Vb accumulates at cell edge………………………………...…………… 56
Figure 2-6. Cell edge localization of myosin Vb depends on motor activity…………...….. 57
Figure 2-7. Cells treated with latrunculin…….…………………………………………...…… 58
Figure 2-8. Single-molecule quantification of myosin Vb motility in live cells………...…… 59
Figure 2-9. Myosin Vb processivity…………………………………………………...……….. 60
Figure 2-10. Rif expression increase dissociation rate of myosin molecules from cortical
actin…………………………………………………………………………………. 61
Figure 2-11. Trajectory length distribution of tethered molecule…………………...………. 62
Figure 2-12. Computer model of barbed-end tethering ……………………………...……... 63
Figure 3-1. Myosin Vb induce filopodia formation…………………………………...………. 64
Figure 3-2. Myosin Vb forms tight focus at the growing filopodia tip…………...…………. 65
Figure 3-3. FRAP experiment of filopodia………………...………………………..…………. 66
Figure 3-4. SptPALM single molecule tracking of myosin Vb within the filopodia...………. 67
Figure 3-5. Myosin Vb competes with other barbed-end-binding proteins for actin
Binding……………………………………………………………………………….. 68
xi ABBREVIATIONS
MT microtubule
F-actin filamentous-actin
IQ amino acid sequence of IQxxxRGxxxR
ATPase adenosine triphosphatase
EM electron microscopy
GTPase guanosine triphosphate
LTP Long term potentiation
AMPA α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid
ER endoplasmic reticulum mGLUR1 metabotrophic GLUR1
ARP2/3 actin-related protein-2/3
ENA/VASP Drosophila Enabled vasodilator- stimulated phosphoprotein
Rif Rho in filopodia
WASP Wiskott–Aldrich Syndrome Protein
PIP2 phosphatidylinositol 4,5- bisphosphate
CP Capping protein
FRAP Fluorescent recovery after photobleaching
TIR Total internal reflection
PALM Photo-activation localization microscope
PAFP Photo-activatible fluorescent protein sptPALM single particle Photo-activation localization microscope
MVID Microvilli inclusion disease
MSD Mean square displacement
xii Chapter 1. Introduction
1.1 Introduction to Myosin Motor Proteins
1.1.1 Structure and Function of Myosins
The myosin superfamily is a large and diverse family of proteins that are capable of binding to actin filaments and generating mechanical forces by hydrolyzing ATP.
Myosins are found in all eukaryotic organisms from yeast to mammals, and are separated into at least 35 classes. Most organisms have multiple myosin genes.
Human, for example, has 37 myosin genes producing a diverse set of myosin proteins7.
Some myosin proteins are believed to function as load-bearing molecules, e.g. muscle myosin, some are believed to function as cargo transport molecules, e.g. myosin V and myosin VI1, yet others are thought to directly affect actin cytoskeleton and cell motility, e.g., myosin IIb in non-muscle cells8,9. The overall spectrum of the functions of all different myosin proteins is far from being understood.
Structurally, most myosins molecules are composed of at least three domains: the N- terminal motor domain, the neck domain and the C-terminal tail domain. The motor domain, also called the head domain, is the domain that defines the myosin superfamily and is most conserved among different myosins. This part of the myosin contains an
ATP binding pocket and the actin-binding sequence that allows myosin to bind to F- actin, which in turn activates the ATPase activity of the motor domain. In contrast, the tail domain features the most class-specific variations. Because the tail domain is often involved in interacting with cargoes, the diversity in the tail domain suggests that
1 different myosin proteins interact with different cargoes in cells10. While all myosins are capable of hydrolyzing ATP to generate mechanical force, for a certain myosins, the forces generated are clearly used to move the myosin along the actin filaments. One of the best-studied “moving” myosins is myosin Va. In vitro experiments had shown that myosin Va is capable of moving continuously along actin filaments toward the barbed end of the actin filament4. Ex vivo studies have shown that myosin Va is able to switch tracks11. The speed of myosin Va varies from 0.312 - 0.63 µm/s and the run length varies from 0.7- 2.1 µm. Since the step size of myosin Va is believed to be 36 nm per each
ATP hydrolysis, it takes 20-60 steps before dissociating from the actin filament. The property of being able to move multiple steps before dissociation is called,
“processivity”13-15.
Biochemically, processivity depends on two properties of the motor proteins: (1) the proteins are dimeric, made of two copies with the heavy chains linked together; and (2) they have high duty ratios. The degree of processivity is closely related to the duty ratio, which defines the time that one head of myosin is bound to actin filament, compared with the time of one ATPase cycle. A processive motor normally has a high duty ratio, as it is less likely for both actin-binding domains to dissociate from the actin filament simultaneously. Conversely, a non-processive motor has a low duty ratio, as it dissociates from the actin filament quickily. A dimeric myosin walks along the filament by alternating the stepping of the leading head (in red) and the trailing head (in blue)
(Fig 1-1). During an ATP hydrolysis cycle, each head of a processive myosin can exist in one of four states: the nucleotide-free state, the ATP-bound state, the ADP-Pi-bound
2 state and the ADP-bound state. When in the ATP-bound state, myosin hydrolyzes ATP quickly (Fig 1-1A). The release of Pi is coupled with the power stroke, which moves the lever arm of myosin (Fig1-1B). The ADP-bound state has the highest affinity for the actin filament. Thus, in the absence of ATP binding, the heads of myosin are tightly bound on the filament (Fig1-1C). The myosin head is detached from the filament upon exchange of ADP to ATP, after which a new ATP cycle initiates at the trailing head.
Normally, the high duty ratio is the result of the ADP release on one head being a slow and rate-limiting step of the ATP cycle. Coordinated ATP cycle between the leading head and the trailing head can further enhance the processivity for high duty-ratio motors such as myosin Va16,17.
1.1.2 In vitro properties of myosin Vb
The myosin V family includes three genes in higher eukaryotes, namely, myo5a, myo5b, and myo5c, encoding the heavy chains of the myosin V proteins1,18,19. Myosin
Va and Vb have 74% sequence homology in the head domain. Both myosin Va and Vb are considered to be high-duty-ratio motors2,20. In contrast, myosin Vc shares only 30% sequence homology with myosin Va and had been found to be a low-duty-cycle motor21.
Myosin Va is the most well-studied among all three myosin Vs. However, myosin Vb is believed to be a closely related protein and shares many similarities to myosin Va.
Both myosin Va and Vb consist of both the heavy chain peptides (~210kDa) and light chains (~17kDa) that serve as regulatory subunits. The heavy chains of myosin Va and
Vb form obligatory dimers through the coiled-coil domain, which is part of the tail
3 domain. Dimerization is believed to be important for the motor processivity to increase the time myosin Vs remaining bound on the actin filament, because as a dimer a myosin molecule can alternate one head with the other head on the actin filament without fully detaching. The neck region contains six tandems repeats of the so-called IQ motifs
(sequence of IQxxxRGxxxR). The IQ motifs are binding sites for calmodulin, the most important myosin V light chain, as well as other calmodulin-like light chains. Each light chain occupies one IQ site, thus the dimeric myosin allows for the binding of, at maximum, six pairs, or twelve, light chains. This long neck also enables myosin V to span a relatively long stepping distance of about 36 nm. The tail sequence starts with a
α-helical coiled-coil sequence, which facilitates dimerization of the heavy chains. The rest of the tail domain is believed to serve as mainly a cargo binding site and has more variations among different members of myosin Vs, suggesting that one of the functional differences between myosin V isoforms may be related to the cargo specificities1.
Differences between myosin Va and Vb were also found in their motility properties.
For example, the gliding assay, in which one measures the gliding speed of actin filament driven by myosin molecules immobilized on a glass slide, showed that the gliding speed of actin can be as fast as 0.6 µm/s when driven by myosin Va and ~0.2
µm/s when driven by myosin Vb. The differences between single-molecule assay and gliding assay are small, which suggests that in both cases, the forces were probably produced by one myosin molecule4,20,22.
4 Both myosin Va and Vb are believed to be regulated by calcium ion concentration.
Electron microscopy and sedimentation studies show that myosin Vs can undergo a change in the conformation as they switch from an active, extended state to an inactive, compact one23. At low Ca2+ concentration, myosin Vs adopt the compact conformation, in which the cargo-binding globular tail domain folds back on the motor head domain.
This contact inhibits the ATP hydrolysis of the motor domain, therefore myosin in this conformation is incapable of processive movements. At high Ca2+ concentration, myosin
V adopts the open conformation. Paradoxically, at very high Ca2+ concentration
(>10mM), myosin V seems to have a reduced motor processivity because of dissociation of calmodulin light chain. Dissociation of camodulin from the motor, therefore, may serve as a brake to stop the run of the motor 24,25. Therefore, moderate increase in Ca 2+ may activate the ATPase of myosin V, whereas larger increase in Ca
2+ may end motor function. Unfortunately, most studies of the effects of Ca2+ on myosin’s actin-activated MgATPase activity and motility are carried out in vitro. In vivo effect of Ca2+ remain to be fully investigated.
1.1.3 In vivo functions of myosin Vb
While myosin Vb and Va share many structural similarities, their in vivo functions diverge as evidenced by genetic analysis. Myosin Vb mutation is the main cause of the microvillus inclusion disease (MVID)26, an extremely rare intestinal disorder in human.
Enterocytes are a class of intestinal cells that layer the inner surface of the small intestines. Enterocytes normally develop a dense array of microvilli, which are filopodia- like actin-rich protrusions needed for efficient absorption of nutrient at the apical surface
5 of the cells. Patient with MVID do not develop microvilli or develop mislocalized microvilli27. Conversely, in cellular models, re-expression of myosin Vb in cells lacking microvilli rescues the phenotype, indicating a direct link between functional myosin Vb and the formation of the microvilli28.
In comparison, mutations in myosin 5a produce very different phenotypes. Mutations in a number of alleles of myo5a in mice cause a dilution in coat color due to defects in melanosome trafficking in mice. The most severe allele, referred to as the dilute-lethal, displays a severe neurological disorder characterized by ataxia, opisthotonus and seizure29. The mice with dilute-lethal mutation develop convulsive limb movement, and die at about 3 weeks of age. Less severe alleles, referred to as the dilute-viral, display only pigment dilution30. Ultrastructural and fluorescent microscopic studies indicate that in Purkinje cells inactivation of myosin Va leads to a lack of smooth ER and inositol
1,4,5-triphosphate receptor (IP3R) in synapses, which causes defects in intracellular
Ca2+ stores1,31. This phenotype is similar to a flailer mouse model where mice express the tail domain of myosin Va32, while maintaining the endogenous myosin Va. The tail domain of myosin Va causes a dominant negative effect by inhibiting endogenous myosin Va.
In human, mutations of MYO5A are associated with Griscelli syndrome type I, Cross- syndrome and Elejalde syndrome, whose characterization is similar to the mutation in mice33. Patients display a pigmentary dilution of the skin and the hair, severe
6 neurological problems including motor developmental delay, mental retardation, seizures, and ataxia34.
1.2 Myosin Vb’s role in intracellular transport
Mammalian cells use molecular motors for precise sorting and trafficking of a wide range of intracellular molecules. There are three classes of molecular motors: myosin, kinesin, and dynein. Myosins utilize actin filament tracks35, whereas kinesin and dynein utilize microtubule (MT) tracks36. Both the filamentous-actin (F-actin) and microtubules
(MTs) are asymmetric filaments. Therefore movement of motor proteins on these tracks allows polarized distribution of the motor molecules because motors have direction specificities on their tracks. Consequently, polarized distributions of the cargoes are established in the cells. Since myosin Vb is believed to be a processive motor, similar to myosin Va, it is not surprisingly that many studies have explored its role in intracellular transport. These researches show strong evidence that myosin Vb takes part in the endosome recycling pathway37-39.
1.2.1 Processivity and in vivo cargo transport
An important indication that myosin Vb plays a role in cargo transport is that it is likely a processive motor. However, even for a processive myosin motor, in order to be able to transport cargoes long distance, it should be able to switch F-actin tracks during the transporting process. This is because, unlike microtubules, which can grow tens of micrometers long, actin filaments in cells are typically not longer than one to two
7 micrometers. The ability to switch F-actin tracks has been demonstrated for myosin Va both in vitro, and in live cells. However, it is still unknown whether myosin Vb is capable of this. Conversely, high single-motor processivity may not be necessary for transport in vivo because each cargo can bind multiple motors. In cells, both fluorescence images and EM studies have shown that motor proteins are clustered on the same vesicular cargo22,40-42. Furthermore, in vitro studies show that the travel distance of a cargo is sensitive to the number of motors that are bound to a cargo. For example, beads moved by multiple kinesins move further than beads moved by a single kinesin, indicating that multiple motors bound to the same cargo may increase the processivity of the transport43. On the other hand, when transport is due to a combination of dynein and kinesin, the direction of the beads can rely on the net force, which implies a tug of war mechanism, Fig 1-2 illustrates such a mechanism in hippocampal neurons44, however the same mechanism should also apply to most mammalian cells. A cargo is loaded with many motor proteins. Therefore, vesicles typically exhibit bi-directional motion as well as pausing behavior, which likely results from interactions between multiple motors on the cargo22,45.
1.2.2 Regulation of myosin Vb cargo binding
Myosin Vb is associated with its cargo through the tail domain. The tail domain does not have direct affinity to the lipids, however it interacts with several adaptor proteins, e.g., Rab proteins, which localize to vesicular compartments46. Rab family small
GTPase are a crucial family of proteins implicated in the regulation of endosomal recycling47,48. Similar interactions have been also identified for myosin Va. For example,
8 in melanocytes, myosin Va-dependent transport of melanosomes is regulated by
Rab27a, which is localized to mature melanosomes. When Rab27a is loaded with GTP,
GTP-bound Rab27a recruits melanophilin, which in turn binds to the tail of the melanocyte-specific spliced form of myosin Va. Similarly, myosin Vb is believed to transport recycling endosomes in many cell types37 by interacting with Rab11a, which localizes to recycling endosomes. Myosin Vb binds either directly to Rab11a or to
Rab11a-FIP2 complex via association with FIP2. Finally, recent reports show that both myosin Va and myosin Vb are involved in transporting AMPA receptor38,39 in neurons38,39. In addition to Rab GTPase, cargo binding can also be regulated by adaptor/scaffolding protein complex49, phosphorylation of tail domain50, and protein degradation51.
1.3 Myosin Vb’s role in filopodia formation
1.3.1 Actin cytoskeleton
Actin is an essential component of the cytoskeleton in eukaryotic cells. The actin cytoskeleton functions in a wide range of cellular processes, including maintainenance of cell morphology, motility, contractility, cell division, endocytosis, vesicular trafficking, signaling pathway52. Actin fall into three broad classes: α, β and γ isoforms. All actin isoforms exist in two states: monomeric, or globular (G-), actin and the polymeric, or filamentous (F-) actin. Each actin molecule can bind an adenine nucleotide (either ATP or ADP). Monemeric actin has dimensions of ~67X40X37Å and molecular mass of about 43kD. G-actin can assemble into filamentous (F-) actin. The rate-limiting step of
9 actin polymerization is the nucleation of the G-actin. Actin dimers are very unstable and dissociate back to monomers quickly. However, when monomers are associated with a trimer that represents the nucleus of polymerization, and further G-actin is rapidly added into the assembled filament. Without an initiation factor, G-actin only polymerizes above a certain concentration, namely the critical concentration. G-actin is mainly associated with ATP. ATP bound G-actin is hydrolyzed to ADP bound G-actin after it is incorporated onto F-actin. The kinetics of ATP hydrolysis and release of its product, including the phosphate group (Pi), have been studied extensively. These studies show that the rate of ATP hydrolysis is 10 times higher than the rate of Pi release. Hydrolysis of ATP bound actin occurs at the growing (barbed) end of the filament, whereas ADP bound actin dissociates from the shrinking (pointed) end of the filament. ADP G-actin dissociates from F-actin at the pointed end, is recycled to ATP G-actin form, thus allowing polymerization at the growing end to occur again53,54.
In cells the mechanism of F-actin polymerization from G-actin is more complicated than in vitro, due to various regulatory factors. There are at least 250 actin binding proteins55, which can alter kinetics of the various steps of actin polymerization. For example, ADF/Cofillin and gelsolin sever F-actin and thus increase G-actin concentration. Profilin activates polymerization of F-actin, therefore it increases F-actin concentration. Arp 2/3 complexes capture preexisting filaments and facilitate filament branching. CapZ binds to the barbed end of the filament and tropomodulin binds to the pointed end of the filament. The assembly of F-actin in vivo is tightly regulated to ensure a desired distribution of the length and the concentration of F-actin. Meanwhile the actin
10 cytoskeleton is highly dynamic. Cells achieve this by maintaining a dynamic equilibrium between actin monomers and actin filaments, enabling the cells to adapt to environmental changes54,56 and regulate cell shapes and cell motility in response. For example, cells are able to move by assembling actin filament at the leading edge of the cell and pushing the membrane forward with structures such as filopodia or lamellipodia52,57 – which are actin structures that coexist in most cells58.
1.3.2 Mechanism of filopodia formation
One working model for filopodia formation is that the actin filaments are derived from the lamellipodial actin network. Lamellipodial actin array is filled with Arp2/3 complex- nucleated filaments. These filamements are usually short as the result of capping protein that cap their barbed end 57,59, but in some conditions, the barbed ends are clustered together by tip complex proteins, including ENA/VASP proteins, mDia2 and myosin X60, which prevent binding of the capping proteins. These filaments are then crosslinked by proteins such as fascin to generate actin bundles, which are necessary for the filopodial architecture. Certain cell types lack highly branched actin meshworks, yet still form filopodia, suggesting that filopodia may be produced by multiple alternative mechanisms59-63. Nevertheless, a few molecules have emerged as the central regulators of filopodia, which are discussed below:
Capping protein (CP). The most critical step in initiating filopodia is to prevent CP binding to F-actin. CP is an α/ß heterodimer (α~36 kDa, ß~32 kDa) that keeps actin filaments short. It is highly localized to lamellipodia and has high affinity (Kd=0.1-1 nM) for actin filaments’ barbed ends64. CP depletion results in inhibition of lamellipodia and
11 increased microspike-like structures suggesting that CP is a negative regulator of filopodia formation58. Factors that prevent CP binding to F-actin, such as Ena/VASP, or
Formin, can often promote elongation of actin filaments and filopodia formation64,65.
Fascin A single actin filament is not mechanically strong enough to resist membrane tension and generate protrusion. Therefore, some form of actin bundling is necessary.
The leading candidate for filament bundling in filopodia is fascin, which is a 55kD monomeric protein that cross-links actin filaments into unipolar and tightly packed bundles. In various cells, fascin localizes to filopodial bundles and is highly expressed in specialized cell that are particularly rich in filopodia, such as neurons and mature dendritic cells. Formation of filopodia like bundles in vitro is also promoted by fascin66.
ENA/VASP has a well- established role in the formation and maintenance of filopodia.
Its mechanism seems to be two fold: First it bundles F-actin and may have an overlapping function with other bundling proteins such as fascin; Second, by binding preferentially to the barbed-end of the F-actin, ENA/VASP protects the barbed ends from binding of the capping proteins, thus displaying an anti-capping effect. The protection of filaments via ENA/VASP is through its poly-Pro domain, which also binds profilin, thus inducing filament polymerization as well as barbed-end clustering65,67.
Small GTPase Small GTPases are important regulators of actin cytoskeleton. Among them, CDC42 is the most well-known for regulating filopodia formation. CDC42 interacts with WASP and N-WASP, together with PIP2, to release the autoinhibition of WASP, which leads to activation of ARP2/3 complex. Another small GTPase, Rif (Rho in filopodia), induces filopodia formation independent of CDC4262. The effect of Rif on
12 filopodia formation is thought to be mediated through the activation of the formin. In vitro studies have shown that formin nucleates linear actin filaments, stimulates actin polymerization, slows filament depolymerization, and protects barbed ends from the capping proteins63,68,69.
1.3.3 Myosin Vb is required for microvilli formation
Filopodia are finger-like membrane protrusions that are thin (0.1-0.3 µm) and long
(>1 µm) and are filled with tight parallel bundles of filamentous actin. Filopodia are initiated and elongated by actin polymerization and stabilized by actin-crosslinking proteins. Fast growing barbed ends of actin filaments are orientated toward the tip of the filopodia. The structure is very dynamic. The protrusions and retractions are the result of a dynamic balance between actin polymerization at the barbed ends of the filaments and the retrograde flow of the actin filament bundle. Filopodia dynamics vary among different mammalian cells types. For example, protrusion rates of filopodia are 0.1 µm/s in B16 F1 melanoma tissue culture cells, 0.04-0.1 µm/s in neuroblastoma cells, and <
0.1 µm/s in primary hippocampal neurons61. The main cellular functions of filopodia are still under debate. However, filopodia are believed to play important roles in cell migration, metastasis, wound healing, adhesion to the extracellular matrix, embryonic development, axonal outgrowth, and dendritic spine growth in neurons61.
A large number of myosin motors have been implicated in the formation of filopodia- type membrane protrusions. Among them, the myosin Vb is specifically required for the development of the small intestinal microvilli, which are filopodia-like actin protrusions that are a few micrometers in length on the apical surfaces of the enterocytes. One
13 example is from “brush border” of small intestinal epithelial cells. The brush border is composed of microvilli protrusions that are tightly packed at a density of ~1000/cell and organized into well–ordered hexagonal patterns. Mutations in MYO5b gene in human are the main cause of the microvillus inclusion disease (MID), which is characterized by the loss of the microvilli in the small intestines28. Furthermore, knock-down of myo5b expression in model epithelial cells in vitro causes the loss of apical filopodia in the cultured cells, indicating a direct functional role of myosin Vb in regulating filopodia formation8,26. The reason why myosin Vb is needed for microvilli formation, however, is not understood.
1.3.4 Other myosins implicated in filopodia formation
There are three basic types of filopodia-type protrusions in human: filopoida, microvilli and stereocilia. These are cylindrical membrane protrusion, filled with bundled arrays of F-actin that are unidirectionally oriented toward the plasma membrane.
“Normal” filopodia is the structure that extends from the motile adherent cells to detect the external environment during migration. Typically each filopodium contains 10-20 actin filaments crosslinked together. Myosin X is believed to play an important role in filopodia formation and transport of Ena/VASP to the tips of filopodia70-72. Microvillus is a specialized type of filopodia in epithelial cells. Beside myosin Vb, both myosin 1d and myosin 7b also localize to microvilli and are important for in formation or maintenance.
Finally, sterocilia are a type of filopodia structure in hair cells in the cochlea and vestibular apparatus. They are organized in 50-100 bundles at the surface of the sensory hair cells, and are necessary for the normal functions of hair cells, i.e., hearing
14 sounds and detecting changes in balance of body. Myosin 15a, 3a and 7a are localized to stereocilia and mutations in these proteins result in defects in stereocilium8.
1.4 Single molecule imaging of Single Myosin Motor in cells.
A common approach to study trafficking of molecules in living cells is to monitor the dynamics of cargos including mitochondria5, recycling endosomes 38,39or other membrane bound vesicles6,31. Although these studies have yielded significant knowledge about the traffic mechanism, they often lack direct information about the specific motor protein. A cargo is tethered by many different families of motor proteins including kinesin, dynein and myosin. A cargo is transported via MT-based motor on the
MT and switches back and forth onto actin-based motor on the actin track, or vice versa. Thus, tracking of cargo is not very useful to deduce how an individual motor contributes to the transport of cargo. Furthermore, motor activities can be highly heterogeneous in the different cellular conditions.
A major innovation of this thesis is to focus on the dynamics of individual motor molecules in living cells, by using a customized photo-activation localization microscope
(PALM) (Fig 2-1). In our experiments, the target molecule is tagged with PAFPs (photo- activatable fluorescent proteins), thus the amount of fluorescent molecules can be controlled by illuminating the cells with blue or UV light. The specific PAFP (Fig 2-2) employed in most of my studies is EosFP. This protein can be switched from green to red emission. Under conditions where a few molecules out of all are activated to emit the red color, individual molecules can be detected and imaged from both living cells
15 and fixed cells. The positions of individual molecules are determined by centroid fitting algorithms. This allows a better spatial resolution (30 nm) than conventional optical imaging (300 nm). This tool is highly useful in studying the dynamics of biological processes by obtaining trajectories of bio-molecules in living cells 73-75.
Another purpose of carrying out single-molecule studies is to understand molecular heterogeneity. The behavior of the individual molecules may be very different even in the same cell. For example, diffusion coefficient (D) may vary from region to region due to differences in lipid domain compositions. The molecules may randomly bind to other immobile or mobile molecular species. Cytoskeletal elements may trap or obstruct the movement of individual molecules76. This information is typically not observable in ensemble measurements, because only the average behavior of the molecules is observed. By measuring the trajectories of individual molecules, we can better elucidate the true dynamics of molecules.
1.5 Thesis hypothesis and objectives
The driving question of this thesis work is “why myosin Vb is required for microvilli formation?” Based on our current understandings of myosin Vb, we can envision two possible mechanisms, which are not necessarily mutually exclusive. Firstly, myosin Vb may be required to transport a filopodia regulator, such as VASP or Fascin, to a specific cellular location, which in turn initiates filopodia growth. Considerable efforts have focused on elucidating the transport properties and identifying cargos of myosin Vb.
While these previous studies had significantly improved our knowledge on myosin Vb,
16 the target cargos for myosin Vb that could be linked to microvilli formation have not been yet fully identified. A second mechanism is that the motor protein serves as a direct regulator of filopodia by affecting actin cytoskeletal dynamics in a cargo- independent manner. A key hypothesis of the study is that the myosin Vb molecules occupy F-actin barbed-ends and therefore prevent F-actin capping on the barbed end of
F-actin. The barbed-end binding is a direct consequence of the motor activity of the myosin Vb, because it is known to be a barbed-end directed motor and therefore is by definition efficient in seeking and converging onto the barbed-ends. One implication of this hypothesis is that similar mechanisms may exist for other myosin motors, which is consistent with the ubiquitousness of myosins in formation of various filopodia-type structures.
In chapter two, I will study the in vivo motility of myosin Vb using single-molecule imaging method. The main purpose is to test whether myosin Vb is a processive motor in cells, and to test whether single molecules of myosin Vb undergo long-range transport. I will characterize the motility properties of myosin Vb including the directionality, the dissociation rate, the run length and the speed.
In chapter three, I will directly investigate the function of myosin Vb in cell. The main question is whether the motor activity alone is sufficient to induce filopodia and if so what the mechanism is. I will employ myosin mutants that are incapable of cargo binding. I will also test whether myosin Vb prevents capping protein binding to F-actin in cells by studying co-localization of myosin vb and capping proteins.
17
Chapter 2. Single Molecule Motility of Myosin Vb Suggests Mechanism of Barbed-end Tethering in vivo.
2.1 Abstract
Previous studies have shown that myosin Va is a processive transporter, but little is known of the processivity of myosin Vb. Structural and kinetic similarity of myosin Vb to myosin Va suggests that myosin Va and myosin Vb have similar functions. However, the role of myosin Vb in cargo transport in cell is poorly understood. Here, we use single-molecule fluorescence techniques in cell to address the in vivo motility of myosin
Vb, by using recombinant Eos FP-myosin Vb construct. We find that the motor molecule can localize and accumulate at the cell edge without the need of continuous long- distance movement of individual molecules. We have also verified this behavior by computational modeling. The results suggest an in vivo model where Myosin V is tethered strongly at the barbed-end of the actin filament, and the tethering is facilitated by processive motility.
2.2 Introduction
Vertebrate myosin Va in vitro studies 4,12 suggest that myosin Va is an efficient cargo transporter in vivo. Due the domain homology of vertebrate Va to vertebrate Vb 19 and biochemical studies 20, myosin Vb is suggested as an efficient cargo transporter in vivo as well. Myosin Vb77 is one of the three members of the class V myosins in vertebrates.
Myosin Vb is comprised of a two identical heavy peptide chains, encoded by the myo5b gene, that form a homodimer and six calmodulin light chains. Each heavy chain
18 contains one motor domain that binds to filamentous actin (F-actin) and hydrolyzes
ATPs, and six IQ motifs that are thought to be binding sites for calmodulin, and a tail domain that interacts with cargoes.
Myosin Vb can adopt so-called “extended” or “closed” conformations. Motor activity is inhibited when myosin Vb is in the “closed” conformation by the tail domain of the protein folding to make contact with the motor domain1. This self-inhibition is thought to be regulated in cells by either calcium ion or interaction with cargo adaptor proteins, such as the Rab small GTPases1,78-80. The in vivo functions of myosin Vb are being actively investigated. A large body of works point to a function related to intracellular organelle transport81-85, consistent with the prediction of it being a processive motor. A number of studies have investigated its effect on recycling of membrane receptor proteins such as transferrin receptor and various G-protein-coupled receptors. In excitatory neurons, myosin Vb colocalizes with and mobilizes AMPA-receptor- containing recycling endosomes in respond to synaptic stimulation39,86. Overall these results suggest that myosin Vb may have overlapping functions with other transport motors such as myosin Va. On the other hand, in epithelial cells the functions of myosin
Vb seem to be more unique. For example, myosin Vb plays a role in regulating epithelial tissue polarization37. In humans, mutation in Myo5b gene is the main cause of the microvillus inclusion disease (MVID)27,28, which is characterized by the lack of microvilli in enterocytes, which are epithelial cells at the inner surface of small intestines.
Previous studies of myosin Vb motility in vivo have been limited to the analysis of motility of intracellular cargoes 84,87,88. Vesicular cargoes are typically associated with
19 multiple types of motor molecules, each in multiple copies87. Therefore, motility properties of myosin Vb itself during cargo transport is difficult to interpret. In this work, we used a photoconvertible fluorescent protein toanalyze the motility of myosin Vb at single-molecule level in live cells. In addition, the tail domain of myosin Vb is deleted to remove the motility driven by other motor proteins bound to a cargo. Our results suggest that majority of myosin Vb molecules are tethered on actin cytoskeleton. This tethering is due to its slow dissociation from the barbed-end of the actin filament and thus, results in accumulation of myosin Vb at the cell edge. Furthermore, mutant analysis suggests that ATP hydrolysis is required for this barbed end tethering mechanism.
2.3 Material and method.
2.3.1 Cloning and Mutagenesis
Myosin Vb full length (FL) region was obtained from pcENA6/V5/HISB/Myo5B FL (a gift from Dr. Ehler) by using enzyme EcoR1 and PmE1. The coding region was cloned into pENTR2B (Invitrogen) by using EcoR1 and EcoRV to generate myosin Vb FL entry clones. To generate myosin Vb HMM entry clone, the myosin Vb FL entry clones were cut with enzyme Xba1 and Xho1 and used Klenow fragment (New England Biolabs,
#M0212S) to retain 3’to 5’ DNA polymerase activity only, and ligated by T4 ligase (New
England Biolabs, #M0202S). Myosin Vb HMM G441A was generated by PCR based site directed mutagenesis from the template of entry myosin Vb clones. The primer used were: 5’-GGCTGTCAGGGGTGAGGCCAACTGGCTACCGGAAGC-3’, 5’-
CCCAGCTTTCTTGTACAAAGTTGGCATTATAAGA-3’ which generate Afe1 and Xba1 restriction sites and the PCR products ligated to myosin Vb HMM backbone cut by Afe1
20 and Xba1 enzyme. Myosin Vb S1 was generated by PCR. The primer used were: 5-
CACCATGACGTACAGCGAGGTCT-3’ and 5’-AGAACGGGCCTCGATCTTGA-3’.
Destination vectors, pDEST-tdEos-C was generated by inserting a Gateway Cassette
(Invitrogen), and the expression fusion constructs were generated by recombining the entry clones with the destination vectors by using Gateway LR clonase (Invitrogen).
2.3.2 Cell culture and transfection
COS7 (Monkey kidney fibroblast) cells were cultured in DMEM medium containing 10 %
FBS and 1% penicillin/streptomycin. COS7 cells were plated on 35 mm glass bottom dish, and transfected with Lipofectamine 2000 (Invitrogen) according to the manufacture’s suggestion 24 hours before imaging.
2.3.3 Drug treatment
Cytochalasin D was purchased from Sigma-Aldrich and used at 5-500 nM or 10 µM.
Latrunculin B was purchased from Sigma-Aldrich and used at 1µM.
2.3.4 Western blotting
COS7 cells were transfected with tdEosMyosin Vb FL or tdEosMyosin Vb HMM with
Lipofectamin 2000 (Invitrogen) for 36 hours before harvesting. Cells were lysed by KLB buffer containing 150mM NaCl, 25mM TrisHCl, 5mM EDTA, 1mM PMSF, 1mM sodium orthovanadate, 1% Triton X-100, 10% glycerol, 0.1% sodium pyrophosphate, 10 mM ß- glycerophosphate, 10 mM sodium fluoride, and 5µg/ml Aprotinin (Sigma A6279).
21 Proteins were separated by 4-15 % SDS/PAGE and transferred to a nitrocellulose membrane. The membrane was incubated with primary antibody against myosin Vb
(mouse monoclonal, Santa cruz, sc-135995) at 1:500 dilution.
Protein bands were visualized by a chemiluminescence detection system.
2.3.5 Single molecule Imaging and analysis
The single molecule microscope is equipped with 60x microscope objective (NA = 1.45,
Olympus) and a TE-cooled EM-CCD camera. For imaging, the culture medium is switched to MEDIUM199 (Invitrogen) containing 2mM HEPES and 2 % FBS or
Hibernate E containing 1% FBS in 37 °C chamber. Morphology are traced according to the green signal (excitation 488 nm laser line from an argon ion laser at a power density of ~0.04 kW/cm) and differential interference contrast (DIC) image. Single molecules were imaged in the red channel (excitation 532 nm DPSS diode laser at a power density of ~0.4 kW/cm) and the light source for the photo-activation is a 405 nm diode laser.
Single molecule time-lapse images were taken at 100 ms time acquisition. Fluorescent spots corresponding to single molecules were detected via intensity threshold. The positions of the particles were obtained by fitting the fluorescence signal with a 2D
Gaussian function. The particle trajectories were tracked over time by minimizing the total displacements of all particles in adjacent frames of 10 frames/sec. The software for the particle tracking analysis is available in public domain and can be downloaded at http://www.ccam.uchc.edu/yu/Software.shtml.
22 2.3.6 MSD analysis
Mean Squre Displacement (MSD) 76,89,90 Dynamics of the single molecules can be determined from MSD. MSD is calculated from the equation below, where xi and yi are the positions of the single molecules in frame I, recorded with a time interval (Δ t). MSD plot versus time can be obtained on linear axes (Fig 2-3). A straight line indicates a diffusive motion (Fig 2-3B) and a parabolic curve indicates directed motion (Fig 2-3A).
On log-log axes, the diffusive exponent (α) was determined from the slope of the relationship between the MSD and time interval (nΔt): for an immobile particle, α=0, for a randomly moving particle α =1, for a particle moving in a straight line at a constant velocity α =2.
2.4 Result.
2.4.1 Myosin Vb localizes to cell edge
To track single molecules of myosin Vb, we made a recombinant murine myo5b construct which tags the N-terminal of the heavy chain of myosin Vb, with a fluorescence protein, tdEos. The fusion construct expresses the chimeric protein in
COS-7 cells. COS-7 cells have the advantage of not expressing endogenous myosin Vb protein (Fig 2-4), therefore we avoid the complication of forming “hybrid dimer” when we express the recombinant myo5b protein in cells. When full length myo5b is expressed,
23 the fluorescence signal distributes throughout the cell body region, suggesting that the molecules predominately localize to cell cytosol (Fig 2-5A). Furthermore FRAP experiment results indicated very high diffusion rate, confirming that the molecules are freely diffusing in cytosol and are not bound to actin cytoskeleton. This is expected, because intracellular myosin molecules typically exist in an inhibitory conformation by folding on itself, blocking ATPase activity and the actin-binding capability of the molecules. It is hypothesized that cargo binding would release the inhibition and activate the motor molecule. However, only a small fraction of the molecules were found localized to vesicular structures at the perinuclear region. Presumably these molecules were bound to plasma membrane in vesicle recycling systems as previously described88.
To allow constitutive activation of myosin molecules, we made a truncated mutant of myosin Vb, which includes only the N-term 1201 residues. This truncated construct mimics the length of the so-called heavy meromyosin (HMM) proteolytic product of myosin Va, and therefore we named it myo5bHMM. Without the C-term cargo binding domain, myo5bHMM does not inhibit the motor activity by the tail-motor folded conformation. Indeed, when myo5bHMM is expressed in COS-7 cells, the fluorescence signal was found to localize to the leading edge of the adherent cells (Fig 2-5A). The localization is very specific to a narrow region of no more than 0.5 µm, as evidenced from the fluorescence intensity linescans (Fig 2-5D). Furthermore, the localization is not uniform along the leading edge, but tends to cluster and often localize to microprotrusions (Fig 2-5B). Finally, other C-term deletion constructs, that have different
24 lengths of the coiled-coil domain sequence, showed similar edge localization as myo5bHMM (Fig 2-6A).
2.4.2 Cell edge localization of myosin Vb depends on motor activity
To test whether the leading edge localization is due to the motor activity of the myosin, we created a G441A point mutation in myo5bHMM construct. Previous studies showed that the mutant can still bind to F-actin with high affinity, but cannot hydrolyze
ATP and thus won’t be able to move on F-actin31. Indeed, we found that myo5bHMM-
G441A mutant lost the property of leading-edge localization (Fig 2-6B). Furthermore, myo5bS1, a construct with all the coiled-coil domain sequence removed and thus cannot form a dimer, also lost the ability to localize to cell edge (Fig 2-6A). Dimer formation is believed to be critical for motor processivity. Therefore, we conclude that only an active processive motor, such as myo5bHMM, is able to migrate to and accumulate at cell edge.
2.4.3 Single molecule quantification of myosin Vb motility in live cells
Single molecule motility of myosin Vb
At first sight, the accumulation of myosin Vb at the cell edge seems to be the exactly what one whould expect for a processive motor. Actin cytoskeleton is polarized near the leading edge. Myosin Vb is a barbed end directed motor. Therefore, the motor activity should continuously transport a flux of motor molecules towards the cell edge, resulting in higher concentration of motor molecule at the cell edge. On the other hand, in vivo
25 motility parameters of myosin Vb, including motor velocity and processivity, are not known.
To understand myosin Vb motility in vivo, we carried out in vivo single-molecule tracking analysis of the molecule using the sptPALM technique91,92. Because tdEos, the fluorescent protein we used to label myosin Vb, is photo-activatable, the amount of chromophore in activated state can be controlled by UV illumination allowing single molecule detection and tracking molecular positions during time-lapse movies. Attempts to detect single molecules of full length myo5b with this strategy resulted in only observation of diffusive signal, consistent with the notion that the full-length protein is in an inhibited conformation that is unlocalized and undergoing very fast diffusion. In contrast, single molecules of myo5bHMM and myo5bHMM-G441A are both easily detected with high signal-to-background ratio (Fig 2-8A). Latrunculin treatment of cells redistributes the binding of myo5bHMM molecules and reduces the total molecules detected (Fig 2-7), indicating that the molecules are primarily bound to F-actin cytoskeleton. However, the time-lapse movies show that only a very small fraction of the motor molecules are undergoing directional movement, which is expected as ATPase motor activity on the filament, while the majority of the molecules are immobile.
Constructs with longer sequences of the coiled-coil domain (myo5b-N1294 and myo5b-
N1630) produce similar results. Hereafter we will refer to the molecules undergoing directional motion as the “moving” molecules and the molecules immobile as the
“tethered” molecules (Fig 2-8B).
26 To quantify the number of moving versus the tethered myo5bHMM molecules, we divided the positional trajectories of each molecule into segments of 300 ms and calculated the spatial displacements of each segment. Figure 2-8C shows quantification of the distribution of the displacements. For the G441A mutant, the apparent displacement is almost completely due to measurement errors. We assume the measurement errors is the combination of two independent random errors from the measurements along the X axis and Y axis, sx and sy, each following a normal
!!!/!!! distribution: �!" � = �!" � = � . Thus the total measurement error is
! ! � = �! + �!
The distribution of S follows the form:
!! � ! � � = � � !!! ! ! �!
Indeed the displacement distribution of myo5bHMM-G441A can be well-fit with this equation (Fig. 2-8C, left), allowing us to determine the value of σ. Once the parameter σ is determined, the distribution of myo5bHMM molecules can be decomposed into two parts (Fig 2-8C, right) – the directional part representing the moving population and the immobile part representing the tethered population – by trying to best match the early segment of the distribution (r < 0.05 µm) with the same equation of Pt(r). Based on this procedure, we estimate that for myo5bHMM, 89.5% of the molecules are tethered, while
10.5% are moving. The average displacement of the moving population is 0.080 µm, corresponding to an average velocity of 0.33 µm/sec, which is the same ordered magnitude, but slightly faster than the actin gliding velocity (0.22 µm/sec) with purified
27 myosin Vb20. The existence of a large fraction of tethered population, however, suggests that the motor can be easily stalled.
2.4.4 Myosin Vb processivity
We also quantified the processive run lengths of the motor molecules by computing the distribution of the end-to-end displacement of each molecule’s trajectory (Fig 2-9A).
In order to exclude molecules that were completely tethered, only trajectories with run length > 0.2 µm were included in the distribution. A distance of 0.2 µm is almost an ordered magnitude greater than the average measurement error and therefore is highly unlikely to be a false positive. The run length distribution can be fit with a single exponential function. Exponential fitting of the distribution suggests that the characteristic run-length, l, of the myo5b motor is 85 ± 19 nm.
The short in vivo run-length of the myosin Vb is also reflected in the mean square displacement (MSD) quantification of the molecule (Fig 2-9B). Most MSD calculations in single particle tracking experiments are carried with combined ensemble and temporal averaging. However, in this case, we suspect that the trajectories have two behaviors: upon initial binding to F-actin, the motor molecule is likely to be in the moving state, while the in later part of the trajectory the molecules are likely to be tethered. Therefore, only ensemble averaging, but not temporal averaging, was carried out. For processive motors, the mean square displacement should scale roughly with square of the time.
However, the experimentally measured MSD for myo5bHMM exhibited sub-linear dependence and a saturation value (Fig 2-9B), somewhat analogous to the case of
28 corralled diffusion93. We hypothesized that this is due to the molecule switching into a tethered state at longer time delay. If this is the case, the MSD function should be approximately:
! !!" ! ! ��� � = �!�� � � �� !
�!�! = � �![1 − + �� + 1 �!!"] ! 2
where l is the characteristic run length, k is the transition rate from the moving state to the tethered state, and γ0 is the probability of starting the trajectory with the moving state. Fitting the experimental data with this equation allows us to estimate the transition rate k ~ 4.3 ± 0.7 sec-1, corresponding to an average run-time of 1/k = 0.23 sec. We can also estimate the average speed of the moving molecules based on this analysis: v = l k
= 0.37 µm/sec, which agrees with the estimation based on the displacement distribution in the previous section.
2.4.5 Rif expression increase dissociation rate of myosin molecules from cortical actin.
Results from single-molecule analysis suggest that there are two populations of myosins. The first population binds to actin far away from barbed-end and dissociate from F-actin quickly. The second are molecules “jammed” at the barbed-ends. To see if this is true in general, we analyzed the trajectories of single motor molecules (Fig 2-10).
We found that the trajectory length distribution contains two exponential components: a fast decaying component with a time constant of 0.45 sec and a slower one with a time
29 constant of 2.4 sec (Fig 2-10, open diamond). Furthermore, the trajectory length distribution with more than 0.5 sec contains one exponential component (Fig 2-11).
Therefore, we hypothesize that if the slow dissociating population is indeed due to molecules bound at the barbed-ends, we should be able to reduce the chance of barbed-end binding by increasing the average length of the actin filaments in the cell.
Formin is an actin nucleation factor that strongly stimulates F-actin extension. Therefore we tested if activating Formin in cells alters the distribution of myosin trajectory lengths.
We found that in cells expressing constitutive Rif, an activator of formin, the trajectory- length of the myo5b molecules are significantly shorter. Interestingly, trajectory length distribution from cells expressing Rif still fits well with two decay component and the time constants (tfast: 0.45 sec and tslow: 2.4 sec) (Fig 2-10, open triangle) are not significantly different from control cells. However, the fraction of the slow component is significantly lower in Rif-expressing cells (14.3%) versus control cells (56.5%). These results suggest that the intrinsic dissociation rates of the non-stalled or the stalled molecules are not altered, but instead the probability of “jamming” is reduced by Rif expression.
2.4.6 Myosin Vb tethering at actin barbed-end
The single-molecule mobility analysis indicates two surprising properties of the motor protein, namely that myosin Vb molecules are predominately tethered with relatively short run lengths. How do we reconcile these observations with the ensemble observation that myo5bHMM accumulates at the cell edge? It is worth noting that both the single-molecule experiments and the ensemble observations were made using the
30 same constructs in the same cell type, thus a proposed mechanism must be consistent with both observations. The first possibility is that myo5bHMM molecules continuously move along the polarized actin cytoskeleton near the cell edge, thus generating a flux of molecules towards cell edge. However, the processive run length of the molecule seems to be too short to counteract the effect of cytosolic diffusion once the motor molecule dissociates from F-actin. The second possibility is that the motor molecules specifically bind to targets that are located at the cell edge. However, this cannot explain why the accumulation of myosin Vb molecules depends on motor ATPase- activity. The third possibility is that myosin Vb tethers to the plus-ends of F-actin. The barbed-ends of F-actin are most concentrated at the cell edge, but are distributed throughout the rest of the actin network. More importantly, this mechanism provides an explanation of why the localization of myo5bHMM at the edge is a motor-activity- dependent: Direct binding of the motor molecule to the barbed-end is inefficient due to the competition of bindings to other regions of the actin filaments. However, due to the motor activity, myosin Vb can move to barbed-end of the F-actin, even if the initial binding is at a random location of the filaments, allowing efficient tethering.
To validate the barbed-end tethering mechanism, we used simple computer models to numerically evaluate whether such a mechanism is sufficient to reproduce the observed cell edge accumulation of the motor molecules, using the motility parameters that we measured in this experiment. The key assumptions of the models are the followings: motor molecules were assumed to exist in three states: moving, tethered and cytosolic diffusion. Once bound to F-actin, a motor molecule is assumed to be in
31 the moving state and starts to move towards the barbed-end of the F-actin. The moving molecule can either dissociate due to the limited run-length, or becomes tethered if it reaches the barbed-end (Fig 2-12A). Finally, we assume that actin filaments are highly polarized near the edge of the cell, but not at the rest of the cell.
We computed the steady state spatial distribution of myosin concentration in a cell using a Monte Carlo method. We found that the tethering mechanism can successfully reproduce the accumulation of motor molecules at the cell edge, without the need for continuous moving the motor molecules on actin cytoskeleton (Fig 2-12B). In comparison, the results of a control model in which the motor molecules dissociate immediately after reaching the barbed-end also resulted in a higher concentration of the motor molecules at cell edge, but the concentration decreases gradually over a shallow gradient through the lamellipodia (Fig 2-12B), which is different from the experimental observation. To quantify the degree of cell-edge accumulation, we used the ratio of molecular concentration at the cell edge and the concentration at one micrometer away from the edge. The ratio is significantly different between the two models (Fig 2-12).
Finally, dependence on motor processivity is investigated by varying motor run-lengths
(Fig 2-12C), and the results indeed show that the efficiency of the edge localization increases as motor run-lengths increases.
2.5 Discussion
In conclusion, by studying the in vivo localization and dynamics of myosin Vb at both the single-molecule and the ensemble level, we found that the motor molecules can localize and accumulate at the cell edge without the need of continuous long-distance
32 movement of individual molecules. Instead, the localization seems to be realized by many cycles of short-range movement, tethering and diffusing to new location. The single molecule motility of myosin Vb is very different from that of myosin Va. Single- particle tracking of myosin Va in cells showed that motors move continuously in a random-walk fashion, indicating that the molecules switch actin tracks every few ATP hydrolysis cycles89. In contrast, myosin Vb molecules seems to be unable to switch actin tracks unless via cytosolic diffusion after they dissociate from the F-actin.
One feature that was not well-reproduced in the numerical simulation is the clustering of the myosin Vb signal at the cell edge, particularly in regions where microprotrusions were formed. In simulation, the myosin was evenly distributed along the cell edge. The clustering likely reflects heterogeneities in the actin cytoskeleton.
More importantly, binding to the barbed-end may affect the polymerization dynamics of the actin filaments, because the barbed-ends are the most important sites for regulating actin polymerization. Most of the barbed-end binding proteins, such as capping protein and the Formin, are well-known regulators of actin cytoskeleton. Therefore, the interplay between myosin Vb and the F-actin may be the source for such heterogeneous binding of myosin.
33 Chapter 3. Myosin Vb Exibits Actin Barbed-end
Binding Activity and Promotes Filopodia Growth.
3.1 Abstract
Many unconventional myosins are believed to be motor molecules directed towards the barbed-end of the actin filaments in cells. However, it is unclear whether in vivo the motor molecules fall off the barbed-ends of F-actin or are “jammed” at the barbed-end.
We investigated cellular distribution, dynamics and single-molecule motility of myosin
Vb, an unconventional family V myosin that is required for microvilli formation in small intestine. We found that the motor activity of myosin Vb is sufficient to promote filopodia formation, and myosin Vb associates stably with the tip of the filopodia. Furthermore, over-expression of myosin Vb compete with other known barbed-end binding protein for localization, suggesting that myosin Vb molecules are “jammed” at actin barbed-end in vivo, and have an effect on actin dynamics. These results support a direct role for myosin in regulating actin cytoskeleton and provide mechanistic insight into myosins’ role in filopodia-like structure formation.
3.2 Introduction
A large number of unconventional myosins are found to be important for the formation of filopodia or filopodia-like cellular structures8. Among them, myosin Vb is specifically required for development of the small intestinal microvilli, which are filopodia-like actin protrusions that are a few micrometers in length produced in a tightly-
34 packed formation on the apical surfaces of the enterocytes. Mutations in MYO5b gene in human are the main cause of the microvillus inclusion disease (MVID), which is characterized by loss of the microvilli in the small intestinal28. Furthermore, knock-down of myo5b expression in epithelial cells causes loss of apical filopodia in cultured cells, indicating a direct functional role of myosin Vb in regulating filopodia formation. Myosin
Vb is comprised of a two identical heavy chains, encoded by the MYO5b gene in human, that homodimerize via the coiled-coil domain. Each heavy chain contains one motor domain that binds to filamentous actin (F-actin) and hydrolyze ATP, and a long neck region containing six IQ motifs that are thought to be binding sites for calmodulin or other light chain peptides. The C-terminal domain of myosin Vb can interact with small GTPases of the Rab family, such as Rab11, either directly or via adaptor proteins such as FIP2, suggesting that myosin Vb also plays a role in intracellular membrane trafficking1.
How myosin Vb is associated with microvilli formation and vesicle trafficking is not understood. Since myosin Vb is likely to play a role in vesicular transport, one mechanism could be that the formation of the microvilli requires specific localization of a protein factor or factors, which the myosin Vb is responsible for the transport of. So far, no such factor has been identified yet. Another possibility is that myosin Vb has a direct effect on the structure or dynamics of the actin cytoskeleton. Very little work has been done in this latter regard. However, for myosin X, another myosin that is a potent filopodia inducer, a truncated molecule without the tail domain is sufficient to induce dynamic formation of microprotrusions94, suggesting that the motor activity itself has an
35 effect on actin organization. One mechanism for myosins to have a direct effect on actin is by binding to the barbed-ends of the F-actin. It is well-understood that the actin barbed-end is a critical site of regulation of actin assembly. Many proteins that regulate filopodia formation, e.g., capping protein, VASP and formin, are barbed-end binding proteins95-97. Meanwhile, most unconventional myosins except for myosin 6 are natural barbed-end seeking molecules due to their motor activity. The question, however, is whether the motor molecule “falls off” at the end of an actin filament once it reaches the barbed-end, or whether it remains associated with the barbed-end in vivo. So far this question has not been addressed in live cells.
In this study we set out to investigate how myosin Vb interacts with actin filaments in cells. We found that the motor domain of myosin Vb can promote filopodia growth in cells. Furthermore we show several lines of evidences from localization study to single- molecule motility assay that support a model of myosin binding to the actin barbed-end.
3.3 Material and methods.
3.3.1 Cloning and constructs.
The Entry clones are generated in chapter 2 Material and Methods.
Destination vectors, pDEST-tdEos-C, pDEST-mEos-C, pDEST-mcherry-C, pDEST-
Cerulean-C were generated by inserting a Gateway Cassette (Invitrogen), and the expression fusion constructs were generated by recombining the entry clones with the destination vectors by using Gateway LR clonase (Invitrogen). Entry clone of
Calmodulin was purchased from Open Biosystems.
36 mRFP-VASP (Dr. Frank Gertler) , GFP-VASP (Dr. Yi Wu) were gifts.
3.3.2 Cell culture and transfection
MCF7 (Human Epithelial cell line), B16F1 (Mouse melanoma cell line), COS7 (Monkey kidney fibroblast) were cultured in DMEM medium containing 10 % FBS and 1% penicillin/streptomycin. Cells were plated on 35mm plastic dishes, transfected with
Lipofectamin 2000 (Invitrogen) according to manufacturer’s recommendation, and grown for 24 hours. MCF7, cells are replated on 35 mm glass bottom dish (Matek) coated with collagen (Invitrogen) 30 µg/ml in 0.02 M acetic acid. B16F1 cells are plated on 35 mm glass bottom dishes coated with laminin (Invitrogen) at 25 µg/ml in PBS for 1 hour. COS7 cells are plated at 35 mm glass bottom dish without coating.
3.3.3 Immunofluorescence and antibodies
Immunostaining was performed after methanol fixation of extracted cells for 5 min with
1% triton X-100 in PBS buffer. For α subunit of CapZ (gift from Dr. Svitkina) staining, primary antibody concentrations was used at 5 µg/ml and secondary antibody
(molecular probe 647 #4414S) was used at 1:1000 dilution. For endogeneous myosin
Vb staining, primary antibody (Rabbit polyclonal, Novus Biogical, NBP1-50217) was used at 1:500 dilution. Secondary antibody (molecular probe 647 #4414S) was used at
1:1000 dilution. Texas red phalloidin (0.033uM, Molecular probes) was used for actin staining.
37 3.3.4 Total internal reflective fluorescent (TIRF) imaging and live cell imaging
Time-lapse live-cell imaging or fixed cell imaging with EPI illumination and TIRF illumination were performed on TIRF microscope (Zeiss) equipped with 100x microscope objective (numerical aperture, 1.45). Cells were imaged in Medium 199 containing 20mM HEPES and 2% FBS or Hibernate E with 2% FBS at RT. For live cell imaging, cells were kept on the microscope stage at the room temperature for 5 mins.
Images of live cells and fixed cells were processed and analyzed using ImageJ and
MetaMorph (Molecular Devices). Intensity quantification was done with ImageJ after background substraction.
Quantification of filopodia density was carried out on fixed cells stained with Texas red phalloidin. Filopodia numbers were counted in regions with prominent leading lamella.
To quantify filopodia tip motility, COS7 and MCF7 were transfected with tdEos-actin or tdEos-Myosin Vb, and DIC and fluorescent images were taken at 5 s interval for 5 mins.
The tip positions from the DIC and fluorescent images were traced manually in ImageJ and the resulting data were analyzed in Excel. Kymographs were generated in ImageJ.
3.3.5 Confocal imaging and fluorescent recovery after photobleaching
(FRAP)
FRAP experiments were done on confocal microscope (Zeiss) equipped with x60, NA
1.4. Live imaging cells were maintained in MEDIUM 199 containing 20mM HEPES and
2% FBS at 37 °C. Cells were transfected with tdEos-Myosin Vb and with no UV
38 activation. The images are collected for 5 mins, with 5s time lapse after 2s photobleaching.
3.4 Results
3.4.1 Myosin Vb induces filopodia formation
To test whether myosin Vb can promote filopodia growth, we over-expressed myo5b protein in MCF7 epithelial cells. Isolated MCF7 cells typically grow apical filopodia near the dome of the cell and rarely produce filopodia at the cell edge around the basal membrane. Therefore, we focused on imaging the basal membrane of the cell using total internal reflection microscopy (Fig 3-1A), which provided a good contrast to test whether the cells had increased filopodia formations. We found that myo5b expression produced increased numbers of filopodia at the cell edge. Fig 3-1 B shows Tirf images from fixed cells that Myosin (tdEos) in green and actin (phalloidin) in red. In some cases, filopodia-like protrusions were found at the ventral membrane, not just at the lateral membrane. Interestingly, a HMM (heavy meromyosin)-like tail-deletion mutant, myo5bΔC, which cannot bind cargo but can still form dimers, is more potent in inducing filopodia than the wild type myo5b, suggesting that filopodia formation is mainly due to the motor activity of the myosin. In either case, the filopodia are highly dynamic and undergo fast protrusion/retraction cycles. To verify that the effect is dependent on the motor activity of the myosin, we created a G441A mutant of myo5bΔC, which is defective in ATP hydrolysis and therefore cannot function as a motor. We found that cells expressing this G441A mutant showed no change in filopodia numbers (Fig 3-1 C).
39 From these experiments we conclude that motor activity of myosin Vb promotes filopodia growth in cells.
3.4.2 Myosin Vb is concentrated at the growing filopodia tip
In many filopodia produced in cells expressing GFP-myo5bΔC, we found that the myosin signal localized to the tip of the protrusion, although some heterogeneity exists.
To further investigate myosin localization, we carried out time-lapse imaging of live cells to see how myosin localization varies over time (Fig 3-2 A). We found that during the protrusion stage of the filopodia, myosin molecules typically localize within a tight focus with close to diffraction-limited spot size at the tip; however, during retraction, the myosin signal becomes more diffuse, possibly because the actin bundle is being severed into shorter fragments during retraction. One filopodium often undergoes multiple rounds of protrusion and retraction cycles, as can be seen from the kymograph of the fluorescence signal (Fig 3-2 B). We found that the filopodia protrusion rate is typically faster than the retraction rate: the average protrusion rate is 0.21±0.016
µm/min, while the average retraction rate is 0.15±0.012 µm/min.
The highly specific tip localization of the myosin signal is particularly striking during protrusion. We measured the intensity distribution along the filopodium during protrusion and found it to be almost identical to the experimentally measured point spread function (PSF) of the microscope (Fig 3-2 C), suggesting that the molecules localize to a very tight focus. Assuming the observed intensity profile is the convolution of the molecular intensity and the PSF, we estimated that the actual dimension of the
40 intensity foci is ~ 63±25 nm, assuming the myosin concentration decays exponentially from the tip. Such a fast decay suggests that the myosin molecules are undergoing very little free diffusion at the tip of the filopodia: assuming a diffusion constant of 5 µm2/sec, it would take ~0.3 ms for the molecule to diffuse out of the 63 nm focus. Therefore it is unlikely that the myosin molecules at the filopodia tip are falling off actin barbed-ends and undergoing “normal” diffusion. They are either “jammed” on the F-actin or have significantly reduced diffusion constant. FRAP experiments (Fig 3-3) confirmed that myosin at the tip of the filopodia did slowly recovers, compared to myosin at the neck of the filopodia which recovers fast.
3.4.3 SptPALM single molecule tracking of myosin Vb within filopodia
Next we tested whether myosin molecules were stably bound at the filopodia tip by imaging single-molecules of myosin in the filopodia (Fig 3-4A). The high concentration of myosin at the tip of the filopodia makes it difficult to resolve individual molecules. To solve this problem, we used a technique, called single-particle tracking PALM
(sptPALM) where combined single-particle tracking with single-molecule localization using photoactivated localization microscopy (PALM).This technique allow us to study dynamics of protein in living cells. The fusion protein of Myosin Vb with photo- activatable fluorescent label mEos was expressed in MCF7 cells. By controlling the level of photo-activation, we were able to limit the number of molecules visible in the single molecule detection channel (Fig 3-4B). Figure 3-4C show a montage of time- lapse images of a filopodium. Most particles shown in the images have intensities
41 consistent with a single fluorescent protein (marked by arrows in yellow and blue). At the tip, occasionally particles with the intensity of approximately two molecules (marked by boxes) are observed, indicating that two molecules are very close to each other. The abrupt appearances and disappearances of the particles is also consistent with single- molecule detection, representing events of actin binding and dissociation respectively.
The average intensity of all frames agrees well with the overall green fluorescence signal as expected. However, the question is whether the higher intensity at the top is mainly due to continuous flux of molecules into the tip region, or mainly due to molecules that have a more stable association to the tip. We found many molecules that appear along the shaft of the filopodium, but most stayed for very short time durations, suggesting quick dissociation. The distances travelled by these molecules at such short durations were in most cases unresolvable at our current resolution (~60 nm), suggesting that flux is not a major factor in bringing the molecules to the tip. However, the average trajectory lengths for molecules near the tip (within 0.5 µm) is 1.46 sec with a standard deviation of 0.54 sec; in comparison, the average trajectory lengths of the molecules away from the tip is only 0.41 sec with a standard deviation of 0.26 sec. This is consistent with the hypothesis that the motor molecules are bound stably at the filopodia tip.
3.4.4 Myosin Vb competes with other barbed-end binding proteins for actin binding.
The next question is whether the myosin Vb molecule at the tip of the filopodia are bound to the actin barbed-ends or to other molecules that also happen to localize at the
42 compartment. To answer this question, we decided to test if myosin Vb competes with other known barbed-end binding proteins. We speculated that if myosin Vb is indeed bound to the barbed-ends, it may prevent other barbed-end binding molecules to localize to the same site and therefore, alter the localization of the actin binding molecules. Vasodilator-stimulated protein (VASP) 98 is an important actin regulator that binds preferably to the barbed-end of F-actin to promote actin polymerization, as well as bundling. When we express VASP-mRFP with CFP-myo5bΔCG441A (Fig 3-5 A, bottom) in the cells, the molecules are localized to the tips of the filopodia, consistent with previous findings. However, when VASP-mRFP is coexpressed with CFP- myo5bΔC (Fig 3-5 A, top), only the myosin molecules, not the VSAP molecules at the tip of the filopodia. The VASP molecules are still localized to the filopodia, but the signal distributed evenly through the shaft of the filopodia protrusion, suggesting that the existing myosin molecules prevented VASP to bind to the barbed-ends.
Next we tested if myosin Vb and actin capping protein, another important barbed-end binding protein, which inhibits actin growth, can localize together. To do that, cells expressing CFP-myo5bΔC (Fig 3-5 B, top) or CFP-myo5bΔCG441A (Fig 3-5 B, bottom) were fixed and immunostained with antibodies against the capping protein. We found that for myo5bΔC, the regions along the cell edge that have higher myosin signal rarely have a strong staining for the capping protein, and vice versa, the regions where there is strong capping protein staining typically show little myosin signal or filopodia formation. Importantly, this negatively colocalized pattern was not observed for the myo5bΔC-G441A mutant, which exhibited a positive co-localization with the capping
43 protein. These results are consistent with the hypothesis that myosin Vb associate directly with the actin barbed-ends and can compete with other barbed-end proteins.
3.5 Discussion
Many unconventional mammalian myosins had been identified to play a role in the formation of filopodia and filopodia-like structures. For example, myosin X is a potent inducer of filopodia and is important for angiogenic signaling process, during which the endothelial tip cells grow numerous filopodia. Myosin IIIa , myosin XV, myosin VI, myosin VIIa and myosin Ic are important for the formation and maintenance or stereocilia. Similarly, myosin Ia, myosin Vb, VII, myosin X and myosin XV and required for brush border microvilli. The mechanisms of each myosin in the filopodia formation process may be different and each myosin may exert multiple independent activities in helping filopodia production. For example, several filopodia myosins, including myosin
VII, myosin X and myosin XV, contain FERM domains (Four.one protein Ezrin Radixin
Moesin), which are considered as an important protein domain for localizing proteins to adhesion or plasmamembrane. Therefore, these molecules may help filopodia to adhere to the extracellular matrix. Some of the myosins interact with other actin modulators that promote filopodia, and thus may play a role in transporting these factors to the sites of filopodia formation. For example, myosin IIIa interacts with espin1 and myosin XVa interacts with whirlin8 . However, these mechanisms do not explain why the myosin motor domain can also affect architecture of the actin cytoskeleton, which was reported for myosin X and now for myosin Vb. Our results suggest that a potential mechanism is via myosin’s stable association with the actin barbed-ends and via competition with other barbed-end binding proteins for regulating actin dynamics.
44 Myosins, by design, are very efficient in localizing to actin barbed-ends, because, unlike normal bimolecular interactions, myosins do not need to directly collide with the barbed- end actin monomers in order for the association reactions to happen, which could be inefficient and subject to competition with other non-barbed-end actin units. Instead, a myosin molecule can bind to F-actin anywhere close to the barbed-end and will reach the barbed-end via the motor activity.
A few reports have demonstrated that myosin Vb can undergo long range transport when bound to cargoes85. In contrast, we found that at individual molecule level, myosin
Vb seems to be a motor with low processivity, suggesting that cargo binding is necessary for long-range transport. Myosin molecules bound to cargoes are different from single molecule in at least two regards: First the same cargo is likely to have multiple motor molecules bound, which may increase the processivity and be advantageous for long range transport. Secondly, the cargoes are significantly larger than single motor molecules and may diffuse slowly enough so dissociated motor can rebind back to F-actin to continue transport. Interestingly, a recent study showed strong evidence that long-range transport of myosin Vb is strongly coupled to actin polymerization. The authors argued that the vesicles the myosin Vb bound to contain actin polymerization factors which polymerize actin filaments at the same time being transported by myosin, which is necessary for long range transport in an cellular environment where most actin filaments are shorter than a micrometer. It is unclear what the exact actin polymerization factors are involved in such a process, however our
45 evidence suggests that myosin Vb itself can serve as an actin polymerization regulator and may mediate the linear actin filament growth.
46 Chapter 4. Summary and Future Directions
4.1 Summary of findings
4.1.1 Individual myosin Vb motor molecules are tethered at the barbed end of the filaments in cells.
To elucidate the function of myosin Vb in vivo, we used a photoconvertible fluorescent protein to measure the motility of myosin Vb at single-molecule level in live cells. In addition, removing the tail domain of myosin Vb has two direct consequences:
1) the motor activity is no longer under regulation of the “folded” or “extended” conformation change; and 2) the motor protein no longer binds to vesicular cargoes.
Therefore, the experimental system ensures single molecule observation and simplifies the dynamics. Our results suggest that the majority of myosin Vb is tethered on the actin cytoskeleton and single molecules of myosin Vb are not capable to switch F-actin tracks, which differentiate it from the closely-related Myosin Va. This tethering is due to its slow dissociation from the barbed-end of the F-actin filament and thus, results in accumulation of myosin Vb at the actin barbed ends. At cellular level, this is reflected by the accumulation of myosin fluorescence signal at the cell edge, where the barbed-ends are highly concentrated. Furthermore, comparison with ATPase mutant of myosin indicates that ATP hydrolysis is required for this barbed end tethering mechanism.
47 4.1.2 Myosin Vb motors directly remodel the actin cytoskeleton by protecting the barbed ends of the actin filament.
Many proteins that regulate filopodia formation, e.g., capping protein, VASP and formin, are barbed-end binding proteins. Meanwhile, most unconventional myosins are natural barbed-end seeking molecules due to their motor activity. The question is whether the barbed-end binding activity of myosin Vb has an effect on filopodia formation. I tested the effect of overexpressing myosin Vb in MCF7 epithelial cells and found that overexpression promote filopodia growth. More importantly, expressing the motor domain of myosin Vb alone is sufficient and, in fact, more potent than the full- length myosin Vb, to promote filopodia-rich growth. I demonstrated that myosin Vb promotes protrusion of filopodia structure by clustering at and tracking the tip of the filopodia. Finally, single molecule assay of myosin Vb in filopodia suggests that as actin filaments grow, myosin Vb does not “follow” the extended barbed-end directly through motor activity. Instead, it dissociates from filament, and rebinds to the new barbed-end.
4.2 Future directions
4.2.1 Single molecule assay of other family-V myosins: myosin Va and
Vc
Most unconventional myosins are barbed-end directed motors. Therefore, the mechanism that we propose here for myosin Vb may apply to some other myosins.
According to their biochemical properties, myosin Va and Vb have 74% sequence homology in the head domain, and the motor properties such as processivity or speed
48 are similar. However a key difference between myosin Va and Vb is that myosin Va is able to switch F-actin tracks efficiently during motility, as shown by previous single- molecule study of myosin Va in COS7 cells, while I showed that myosin Vb is incapable of this. It seems reasonable to hypothesize that track-switching reduced the chance of barbed-end binding, in other words, myosin Vb binds on the different actin filaments before completion of walking toward the barbed end of the actin filaments. This might be a key feature to determine whether the motor function is mainly for transport or for actin regulation. To generalize this idea and to understand the general functions of myosin V in cell, we should extend the current study to other myosins, such as other class V myosins: myosin Va and myosin Vc.
4.2.2 Cooperation between myosin Vb and other filopodia regulators
Many other molecules such as the Diaphanous-related formin, Dia2, Ena/VASP, fascin, or myosin X induced Filopodia. The process of filopodia formation is a multi-step process and involves at the very least the process of actin filament nucleation and elongation, barbed-end protection, and filament bundling. While we showed that one reason myosin Vb promote filopodia is through barbed-end protection, it is possible that other mechanisms are also involved. For example, myosin may be involved in the process of filament clustering by binding to two filaments at the same time because each protein has two heads. Furthermore, myosin Vb may interact with other filopodia regulators and recruit these factors to the tip of the filopodia. These alternative mechanisms should be examined in detail in the future experiments.
49
FIGURES
Figure 1-1. Models for myosin V processivity. (A) ATP hydrolyzes immediately after binding to myosin head. (B) Hydrolysis, Power stroke, Release of Pi from the head is coupled with rotation of the lever arm (power stroke). ADP bound myosin head is tightly bound to f-actin. (C) ADP releases. ATP subsequently binds to the nucleotide free head and ATP bound myosin head loosely bound to f-actin. High affinity is when myosin is either in a nucleotide-free or ADP bound head. Low affinity is when myosin is either in a ATP bound or ADP-Pi bound head.
50
Figure 1-2. Tug-of-war mechanism. (A) In cells, a cargo is tethered by many motors: Myosin (in yellow), Kinesin (in red), and Dynein (in green). (B) Orientation of MT is heterogeneous in a compartment, Kinesin can drive the cargo toward plus end of the MT, while Dynein can drive the cargo toward minus end of the MT. (C) Cells have a different concentration of F-actin distribution. Actin rich area plays an important role by myosin motors.
51
Figure 2-1. PALM microscope setup. The photoactivation laser excites the sample under TIR condition. The fluorescence excitation laser illuminates the whold depth of the sample. The unactivated molecules are collected in green; the activated molecules are collected in red. The images are taken every 100 ms.
52 Before UV After UV
Figure 2-2. Normalized excitation and emission spectra of Eos Fluorescent Protein (FP). The Green form: the excitation spectra (black) at the peak wavelength of 490 nm and the emission spectra (green) at the peak wavelength of 520 nm of wtEosFP before irradiation of Ultraviolet (UV). The Red form: The excitation spectra (gray) at the peak wavelength of 560 nm and the emission spectra (red) at the peak wavelength of 590 nm of wtEosFP after irradiation of Ultraviolet (UV).
53
Figure 2-3. Expected mode of motion of myosin V. Example trajectories (left) and MSD on both linear axes (middle) and log-log axes (right): A) processive, The directed myosin Va has a slope of 2 in log-log scales, indicative of an active process. The trajectories of myosin show processive. B) Non- processive, The diffusive myosin Va has a slope of 1 in log-log scales, indicative of an passive process. The trajectories of myosin show a random walk.
54
Figure 2-4. Western blot of myosin Vb in COS7 cell. Western blot shows that myosin Vb is not expressed endogenously in COS7 cell, but can be efficiently expressed exogenously.
55
Figure 2-5. Myosin Vb accumulates at cell edge. (A) Expressing fluorescence protein tagged full-length (right, Myo5bFL) or tail-deletion mutant (left, Myo5bHMM) of myosin Vb in COS7 fibroblast yielded different localization patterns. Full length myo5b localized primarily to cytosol, and vesicular membrane compartments concentrated in perinuclear region (arrows), while myo5bHMM accumulated and localized to cell edge, as well as the tip of filopodia/ micro-protrusions (arrow heads). (B) Dynamics of the cell peripheral localized myosin Vb. Time lapse image series were obtained from COS7 cells expressing myo5bHMM, and plotted in 1-min interval. The myosin Vb fluorescence signal foci can be observed to persist for several minutes and moving along the cell edge (yellow arrow head), or follow the tip of micro-protrusions (blue arrow head), before dissipation. (C) TIRF fluorescence image colocalization of myosin and actin in cells expressing myo5bHMM. Images show myosin (green) localize to regions of actin network, but not to stress fibers. (D) Quantification of the fluorescence intensity shows sharp decrease of myosin signal within the first micrometer distance from the cell edge. Scale bars represent 5 µm unless specified otherwise.
56
Figure 2-6. Cell edge localization of myosin Vb depends on motor activity. (A) Spatial distribution of various myo5b truncation mutants (left) near cell edge. Mutants with full or partial coiled-coil domains all accumulate at the cell edge. The S1 mutant that completely lacks the coiled-coil domain does not accumulate at the cell edge, similar to full length myo5b. (B) TIRF image of the myo5bHMM and a corresponding ATPase mutant (myo5bHMM-G441A). The mutant shows no sign of cell-edge localization and accumulation. Scale bars represent 5 µm.
57
Figure 2-7. Cells treated with Latrunculin. Latrunculin treatment redistribute binding of myo5bHMM. The signal in the region of cell protrusion is significantly reduced while the signal near nucleus region increases. Scale bars represent 3 µm.
58
Figure 2-8. Single molecule quantification of myosin Vb motility in live cells. (A) Representative single molecule images (left) and trajectories (right) of myo5bHMM and the corresponding G441A mutant. Only molecules tracked for more than one second are shown in the trajectory plots so that immobile and moving molecules can be distinguished visually. For myo5bHMM, molecules were either immobile (arrows) or undergoing directional movements (arrow heads). For the G441A mutant, only immobile molecules were observed. Scale bar: 5 µm. (B) Trajectory examples of moving (left) and tethered (right) myo5bHMM molecules. Scale bars represent 2 µm (C) Apparent displacement distribution of myo5bHMM (right) and the G441A mutant (left). Displacements were measured over a time span of 480 ms. The distribution for G441A mutant, which has no motor activity, can be well-fit with equation 1. In contrast, the distribution for myo5bHMM has a significantly heavier tail and does not fit well with the same distribution, but can be decomposed into the tethered component (dotted line) and the residue component (solid line) representing the moving molecules. See text for more details.
59 (A) (B) my o5bHMM myo5bHMM-G441A
0.007
0.006
0.005
0.004 2) m0.003 µ ( SD 0.002 M 0.001
0 0 0.5 1 1.5 2 2.5 3 -0.001
-0.002 Time (sec)
Figure 2-9. Myosin Vb processivity. (A) Run lengths of myo5bHMM molecules. Only molecules with run lengths > 0.2 µm were included in order to remove the immobile fraction. The solid line represents the best single-exponential fit to the data. (B) The mean-square-displacement (MSD) measurement for myo5bHMM and the corresponding G441A mutant. The MSD values of G441A mutant are time-independent, indicating the molecules are immobile. The MSD of myo5bHMM increases with time delay, but is sublinear, indicating that the molecules cannot move persistently over a long time. The dotted line represents the expected MSD value due solely to the measurement error.
60 450
400 - Rif + Rif 350
300
250
200
150 Number of Molecules
100
50
0 0 1 2 3 4 5 Time (sec)
Figure 2-10. Rif expression increase dissociation rate of myosin molecules from cortical actin. The dwell times of single myo5bΔC molecule remain bound at cortical actin layer were measured via sptPALM with TIR illumination. The histograms of the dwell times plotted and fitted with double-exponential model to obtain dissociation time constant and relative population of the two components. In cells expressing constitutive active Rif, the relative fraction of the slow-dissociating population is significantly reduced.
61 120
100
80
60 Counts
40
20
0 0 5 10 15 20 Time (sec)
Figure 2-11. Trajectory length distribution of tethered molecules. Single particle tracing data was obtained from COS7 cells expressing myo5bHMM. Only molecules tracked for more than 0.5 sec were analyzed in order to test whether the molecule is in tethered state. The exponential fit (solid line) to the distribution yielded a time constant of 1.98 ± 0.20 sec.
62 (A) (B) (C)
N o te thering c1 3.0 ) Barbed-endte th ering 1.0 N o tetherin g Fast au ( n 2.5 io 0.8 F-actin at c1 tr / n 0.6 1 2.0 - end + end ce n o 0.4 C Barbed-end tetherin g 1.5 Slow ve 0.2 ti la e 0 1.0 F-actin R -2 -1 0 0 1 2 3 4 5 10 10 10 - end + end Distance from cell edge (µm) Run Length (µm)
Figure 2-12. Computer model of barbed-end tethering. (A) A cartoon depicting the motility model of the myosin Vb on F-actin. The motor molecules bind to F-actin via diffusion in cytosol. If the binding site is far away from the barbed end, the short processive run of the motor does not bring the motor to the end and the motor molecule dissociates. In contrast, if the motor molecule binds close to the barbed end, the processive run will result in motor molecules stably associated to the barbed end. (B) Quantification of the myosin concentration as a function of distance to the cell edge. Simulation results based on the barbed-end tethering model is plotted in solid squares. In comparison, the result based on the assumption that myosin Vb falls off from F-actin once the molecule runs to the barbed ends were shown in solid triangles. Run length of the motor is assumed to be 85 nm in both models. (C) Quantification of the accumulation of myosin at the cell edge, calculated by taking a ratio between the image intensity at the cell edge and the intensity at 1 µm away from the cell edge. The ratio is computed as a function of run length of the motor molecules.
63
Figure 3-1. Myosin Vb induce filopodia formation. (A) Total internal reflection microscopy images of MCF7 cells expressing full length myo5b or mutants. Bottom panels correspond to cell region marked by the white box in the top. Scale bar is 3 µm. (B) Tail domain of myosin Vb is dispensable for inducing filopodia. MCF7 cells expressing full length myosin Vb or mutants were counter stained with phalloidin to label the filopodia. Cells expressing tailless myo5b, but not cells expressing G441A mutant, grew more filopodia than cells expressing full length myosin. Scale bar is 3 µm. Control cells were cells expressing empty vector. (C) Quantification of basal filopodia density of MCF7 cells expressing myosin Vb.
64 (A) (C) 0 s 10 s 20 s 30 s 40 s
1 Protrusion
Protruding Retraction
50 s 60 s 70 s 80 s 90 s 0.8 PSF
0.6 Retracting
(B) 0.4
Myo5bΔC Relative Intensity (au) Tip 0.2
Base 0 2 µm 0 500 1000 1500 2000 4 min Distance (nm)
Figure 3-2. Myosin Vb forms tight focus at the growing filopodia tip. (A) Time lapse image series of protruding (top) and retracting (bottom) filopodia. Cells were expressing GFP-myo5bΔC. Myosin molecules localize at the tip of the filopodia to form a tight focus during the protrusion stage but not at the retraction stage. (B) Fluorescence kymograph of the myo5bΔC signal at a filopodium. The filopodium protruded and retracted multiple times at the same location, always exhibiting tight focus during protrusion and more dispersed signal during retracting. (C) Intensity linescan quantification of myo5bΔC signals within the filopodia. For comparison, the experimentally measured point-spread function of the microscope (dotted line) is also plotted. The intensity profile of myosin during filopodia protrusion is very close to the point-spread function.
65
Figure 3-3. FRAP experiment of myosin Vb. (A)Time lapse image series after photobleaching on the neck of filopodia (top) and the head of the filopodia (bottom). The bleaching area is marked by the rectangle and bleached for 10 seconds. The images are taken every 5 s. scale bar 5 µm. (B) Data were shown for prebleach image at 0s, bleach image at 5s, and post bleach at every 5 seconds for 50s. The intensity pfofiles for each image (frame 1-10) to calculate the FRAP curves for shaft (rectangle) and tip (round).
66
(A) (B)
(C)
Figure 3-4. SptPALM single molecule tracking of myosin Vb within the filopodia. (A) A region of the cell expressing mEos-myo5bΔC. The image was taken with green fluorescence filter to show overall fluorescence intensity distribution of all myosin molecules. The right panel shows a zoomed-in view of one filopodium marked by the white box. Scale bar is 3 µm. (B) Total fluorescence intensity from a time-lapse series of single-molecule images shown in C. The higher intensity at the tip of the filopodia is consistent with the green fluorescence signal shown in A, and indicative of a higher probability of finding single molecules at the tip region. (C) Time lapse series of single myosin Vb molecules in the filopodium shown in A. The images were taken with a red fluorescence filter to ensure only photo-activated mEos molecules were detected. Images were taken every 120 ms and time-series was laid out from left to right and then from top to bottom. Most molecules were detected for only a few hundred milliseconds except at the tip region.
67 (A) Myosin VASP Myosin/VASP C Δ Myo5b G441A - C Δ Myo5b
(B) Myosin Capping Myosin/Capping C Δ Myo5b G441A - C Δ Myo5b
68
Figure 3-5. Myosin Vb competes with other barbed-end-binding proteins for actin binding. (A) Myosin Vb displaces VASP from the filopodia tip. In cells expressing both CFP-myo5bΔC and mCherry-VASP, only myosin signal were found at the tip of filopodia (TOP). In contrast, VASP signal is localized to filopodia tip where the G441A myosin mutant was expressed (Bottom). Scale bars represent 5 µm. (B) Myosin Vb and capping protein occupies non-overlapping regions at the cell peripheral. Cells were transfected to express CFP-myo5bΔC and then immunostained with an antibody against the capping proteins. Regions with high myosin signal exhibited low capping proteins staining and, vice versa, regions with high capping protein signal had little amount of myosin (Top). In contrast, capping protein colocalizes well with G441A mutant of myosin, which presumably binds to F-actin at random location (bottom). (C) Co- localization scatter plot. Left: co-localization plot of Myosin Vb HMM and capping protein. Right: co-localization plot of Myosin Vb HMM G441A and capping protein.
69
Table 1. Model parameters
Parameter Physical meaning Value Ref Actin Tracks 99 Lactin Average filament length 1.4 µm d Size of polarized region 3.0 µm 99 State Switching Kinetics k+ Rate of motor binding to F-actin 25 /sec - k- Dissociation from tethered state 0.5 /sec - l Run length 85 nm - Motility parameters v Motor velocity 0.4 µm/sec - D Cytosolic diffusion constant 10 µm2/sec 100
70 Reference
1 Hammer, J. A., 3rd & Sellers, J. R. Walking to work: roles for class V myosins as cargo
transporters. Nat Rev Mol Cell Biol 13, 13-26, doi:10.1038/nrm3248 (2012).
2 Cheney, R. E. et al. Brain myosin-V is a two-headed unconventional myosin with
motor activity. Cell 75, 13-23 (1993).
3 Wolenski, J. S., Cheney, R. E., Mooseker, M. S. & Forscher, P. In vitro motility of
immunoadsorbed brain myosin-V using a Limulus acrosomal process and optical
tweezer-based assay. J Cell Sci 108 ( Pt 4), 1489-1496 (1995).
4 Yildiz, A. et al. Myosin V walks hand-over-hand: single fluorophore imaging with 1.5-
nm localization. Science 300, 2061-2065, doi:10.1126/science.1084398 (2003).
5 Pathak, D., Sepp, K. J. & Hollenbeck, P. J. Evidence that myosin activity opposes
microtubule-based axonal transport of mitochondria. J Neurosci 30, 8984-8992,
doi:10.1523/JNEUROSCI.1621-10.2010 (2010).
6 Song, A. H. et al. A selective filter for cytoplasmic transport at the axon initial
segment. Cell 136, 1148-1160, doi:10.1016/j.cell.2009.01.016 (2009).
7 Odronitz, F. & Kollmar, M. Drawing the tree of eukaryotic life based on the analysis
of 2,269 manually annotated myosins from 328 species. Genome Biol 8, R196,
doi:10.1186/gb-2007-8-9-r196 (2007).
8 Nambiar, R., McConnell, R. E. & Tyska, M. J. Myosin motor function: the ins and outs
of actin-based membrane protrusions. Cell Mol Life Sci 67, 1239-1254,
doi:10.1007/s00018-009-0254-5 (2010).
71 9 O'Connell, C. B., Tyska, M. J. & Mooseker, M. S. Myosin at work: motor adaptations for
a variety of cellular functions. Biochim Biophys Acta 1773, 615-630,
doi:10.1016/j.bbamcr.2006.06.012 (2007).
10 Krendel, M. & Mooseker, M. S. Myosins: tails (and heads) of functional diversity.
Physiology (Bethesda) 20, 239-251, doi:10.1152/physiol.00014.2005 (2005).
11 Brawley, C. M. & Rock, R. S. Unconventional myosin traffic in cells reveals a selective
actin cytoskeleton. Proc Natl Acad Sci U S A 106, 9685-9690,
doi:10.1073/pnas.0810451106 (2009).
12 Mehta, A. D. et al. Myosin-V is a processive actin-based motor. Nature 400, 590-593,
doi:10.1038/23072 (1999).
13 Baker, J. E. et al. Myosin V processivity: multiple kinetic pathways for head-to-head
coordination. Proc Natl Acad Sci U S A 101, 5542-5546,
doi:10.1073/pnas.0307247101 (2004).
14 Veigel, C., Schmitz, S., Wang, F. & Sellers, J. R. Load-dependent kinetics of myosin-V
can explain its high processivity. Nat Cell Biol 7, 861-869, doi:10.1038/ncb1287
(2005).
15 Sakamoto, T. et al. Neck length and processivity of myosin V. J Biol Chem 278,
29201-29207, doi:10.1074/jbc.M303662200 (2003).
16 De La Cruz, E. M., Wells, A. L., Rosenfeld, S. S., Ostap, E. M. & Sweeney, H. L. The
kinetic mechanism of myosin V. Proc Natl Acad Sci U S A 96, 13726-13731 (1999).
17 Rosenfeld, S. S. & Sweeney, H. L. A model of myosin V processivity. J Biol Chem 279,
40100-40111, doi:10.1074/jbc.M402583200 (2004).
18 Sellers, J. R. Myosins: a diverse superfamily. Biochim Biophys Acta 1496, 3-22 (2000).
72 19 Reck-Peterson, S. L., Provance, D. W., Jr., Mooseker, M. S. & Mercer, J. A. Class V
myosins. Biochim Biophys Acta 1496, 36-51 (2000).
20 Watanabe, S., Mabuchi, K., Ikebe, R. & Ikebe, M. Mechanoenzymatic characterization
of human myosin Vb. Biochemistry 45, 2729-2738, doi:10.1021/bi051682b (2006).
21 Watanabe, S. et al. Human myosin Vc is a low duty ratio nonprocessive motor. J Biol
Chem 283, 10581-10592, doi:10.1074/jbc.M707657200 (2008).
22 Gross, S. P., Vershinin, M. & Shubeita, G. T. Cargo transport: two motors are
sometimes better than one. Curr Biol 17, R478-486, doi:10.1016/j.cub.2007.04.025
(2007).
23 Li, X. D., Jung, H. S., Mabuchi, K., Craig, R. & Ikebe, M. The globular tail domain of
myosin Va functions as an inhibitor of the myosin Va motor. J Biol Chem 281, 21789-
21798, doi:10.1074/jbc.M602957200 (2006).
24 Trybus, K. M. et al. Effect of calcium on calmodulin bound to the IQ motifs of myosin
V. The Journal of biological chemistry 282, 23316-23325,
doi:10.1074/jbc.M701636200 (2007).
25 Coluccio, L. M. Myosins, a superfamily of molecular motors. 7 (2007).
26 Muller, T. et al. MYO5B mutations cause microvillus inclusion disease and disrupt
epithelial cell polarity. Nat Genet 40, 1163-1165, doi:10.1038/ng.225 (2008).
27 Ruemmele, F. M. et al. Loss-of-function of MYO5B is the main cause of microvillus
inclusion disease: 15 novel mutations and a CaCo-2 RNAi cell model. Hum Mutat 31,
544-551, doi:10.1002/humu.21224 (2010).
73 28 Szperl, A. M. et al. Functional characterization of mutations in the myosin Vb gene
associated with microvillus inclusion disease. J Pediatr Gastroenterol Nutr 52, 307-
313, doi:10.1097/MPG.0b013e3181eea177 (2011).
29 Mercer, J. A., Seperack, P. K., Strobel, M. C., Copeland, N. G. & Jenkins, N. A. Novel
myosin heavy chain encoded by murine dilute coat colour locus. Nature 349, 709-
713, doi:10.1038/349709a0 (1991).
30 Seperack, P. K., Mercer, J. A., Strobel, M. C., Copeland, N. G. & Jenkins, N. A. Retroviral
sequences located within an intron of the dilute gene alter dilute expression in a
tissue-specific manner. EMBO J 14, 2326-2332 (1995).
31 Wagner, W., Brenowitz, S. D. & Hammer, J. A., 3rd. Myosin-Va transports the
endoplasmic reticulum into the dendritic spines of Purkinje neurons. Nat Cell Biol
13, 40-48, doi:10.1038/ncb2132 (2011).
32 Jones, J. M. et al. The mouse neurological mutant flailer expresses a novel hybrid
gene derived by exon shuffling between Gnb5 and Myo5a. Hum Mol Genet 9, 821-
828 (2000).
33 Sanal, O. et al. Griscelli disease: genotype-phenotype correlation in an array of
clinical heterogeneity. J Clin Immunol 22, 237-243 (2002).
34 Sanal, O. et al. An allelic variant of Griscelli disease: presentation with severe
hypotonia, mental-motor retardation, and hypopigmentation consistent with
Elejalde syndrome (neuroectodermal melanolysosomal disorder). J Neurol 247,
570-572 (2000).
35 Rayment, I. et al. Three-dimensional structure of myosin subfragment-1: a molecular
motor. Science 261, 50-58 (1993).
74 36 Vale, R. D. The molecular motor toolbox for intracellular transport. Cell 112, 467-
480 (2003).
37 Roland, J. T. et al. Rab GTPase-Myo5B complexes control membrane recycling and
epithelial polarization. Proc Natl Acad Sci U S A 108, 2789-2794,
doi:10.1073/pnas.1010754108 (2011).
38 Correia, S. S. et al. Motor protein-dependent transport of AMPA receptors into spines
during long-term potentiation. Nat Neurosci 11, 457-466, doi:10.1038/nn2063
(2008).
39 Wang, Z. et al. Myosin Vb mobilizes recycling endosomes and AMPA receptors for
postsynaptic plasticity. Cell 135, 535-548, doi:10.1016/j.cell.2008.09.057 (2008).
40 Tabb, J. S., Molyneaux, B. J., Cohen, D. L., Kuznetsov, S. A. & Langford, G. M. Transport
of ER vesicles on actin filaments in neurons by myosin V. J Cell Sci 111 ( Pt 21),
3221-3234 (1998).
41 DePina, A. S. & Langford, G. M. Vesicle transport: the role of actin filaments and
myosin motors. Microsc Res Tech 47, 93-106, doi:10.1002/(SICI)1097-
0029(19991015)47:2<93::AID-JEMT2>3.0.CO;2-P (1999).
42 Encalada, S. E., Szpankowski, L., Xia, C. H. & Goldstein, L. S. Stable kinesin and dynein
assemblies drive the axonal transport of mammalian prion protein vesicles. Cell 144,
551-565, doi:10.1016/j.cell.2011.01.021 (2011).
43 Block, S. M., Goldstein, L. S. & Schnapp, B. J. Bead movement by single kinesin
molecules studied with optical tweezers. Nature 348, 348-352,
doi:10.1038/348348a0 (1990).
75 44 Hirokawa, N. & Takemura, R. Molecular motors in neuronal development,
intracellular transport and diseases. Curr Opin Neurobiol 14, 564-573,
doi:10.1016/j.conb.2004.08.011 (2004).
45 Schroeder, H. W., 3rd, Mitchell, C., Shuman, H., Holzbaur, E. L. & Goldman, Y. E. Motor
number controls cargo switching at actin-microtubule intersections in vitro. Curr
Biol 20, 687-696, doi:10.1016/j.cub.2010.03.024 (2010).
46 Li, D., Miller, M. & Chantler, P. D. Association of a cellular myosin II with anionic
phospholipids and the neuronal plasma membrane. Proc Natl Acad Sci U S A 91, 853-
857 (1994).
47 Seabra, M. C., Mules, E. H. & Hume, A. N. Rab GTPases, intracellular traffic and
disease. Trends Mol Med 8, 23-30 (2002).
48 Seabra, M. C. & Coudrier, E. Rab GTPases and myosin motors in organelle motility.
Traffic 5, 393-399, doi:10.1111/j.1398-9219.2004.00190.x (2004).
49 Kim, E. & Sheng, M. PDZ domain proteins of synapses. Nat Rev Neurosci 5, 771-781,
doi:10.1038/nrn1517 (2004).
50 Yoshimura, A. et al. Myosin-Va facilitates the accumulation of mRNA/protein
complex in dendritic spines. Curr Biol 16, 2345-2351,
doi:10.1016/j.cub.2006.10.024 (2006).
51 Fukuda, M. & Itoh, T. Slac2-a/melanophilin contains multiple PEST-like sequences
that are highly sensitive to proteolysis. J Biol Chem 279, 22314-22321,
doi:10.1074/jbc.401791200 (2004).
52 Pollard, T. D. & Cooper, J. A. Actin, a central player in cell shape and movement.
Science 326, 1208-1212, doi:10.1126/science.1175862 (2009).
76 53 dos Remedios, C. G. et al. Actin binding proteins: regulation of cytoskeletal
microfilaments. Physiol Rev 83, 433-473, doi:10.1152/physrev.00026.2002 (2003).
54 Cooper, J. A. & Schafer, D. A. Control of actin assembly and disassembly at filament
ends. Curr Opin Cell Biol 12, 97-103 (2000).
55 Winder, S. J. & Ayscough, K. R. Actin-binding proteins. J Cell Sci 118, 651-654,
doi:10.1242/jcs.01670 (2005).
56 Pollard, T. D. Regulation of actin filament assembly by Arp2/3 complex and formins.
Annu Rev Biophys Biomol Struct 36, 451-477,
doi:10.1146/annurev.biophys.35.040405.101936 (2007).
57 Nicholson-Dykstra, S., Higgs, H. N. & Harris, E. S. Actin dynamics: growth from
dendritic branches. Curr Biol 15, R346-357, doi:10.1016/j.cub.2005.04.029 (2005).
58 Mejillano, M. R. et al. Lamellipodial versus filopodial mode of the actin
nanomachinery: pivotal role of the filament barbed end. Cell 118, 363-373,
doi:10.1016/j.cell.2004.07.019 (2004).
59 Svitkina, T. M. et al. Mechanism of filopodia initiation by reorganization of a
dendritic network. J Cell Biol 160, 409-421, doi:10.1083/jcb.200210174 (2003).
60 Faix, J. & Rottner, K. The making of filopodia. Curr Opin Cell Biol 18, 18-25,
doi:10.1016/j.ceb.2005.11.002 (2006).
61 Gupton, S. L. & Gertler, F. B. Filopodia: the fingers that do the walking. Sci STKE 2007,
re5, doi:10.1126/stke.4002007re5 (2007).
62 Mattila, P. K. & Lappalainen, P. Filopodia: molecular architecture and cellular
functions. Nat Rev Mol Cell Biol 9, 446-454, doi:10.1038/nrm2406 (2008).
77 63 Rotty, J. D., Wu, C. & Bear, J. E. New insights into the regulation and cellular functions
of the ARP2/3 complex. Nat Rev Mol Cell Biol 14, 7-12, doi:10.1038/nrm3492
(2013).
64 Wear, M. A. & Cooper, J. A. Capping protein: new insights into mechanism and
regulation. Trends Biochem Sci 29, 418-428, doi:10.1016/j.tibs.2004.06.003 (2004).
65 Bear, J. E. et al. Antagonism between Ena/VASP proteins and actin filament capping
regulates fibroblast motility. Cell 109, 509-521 (2002).
66 Vignjevic, D. et al. Role of fascin in filopodial protrusion. J Cell Biol 174, 863-875,
doi:10.1083/jcb.200603013 (2006).
67 Applewhite, D. A. et al. Ena/VASP proteins have an anti-capping independent
function in filopodia formation. Mol Biol Cell 18, 2579-2591, doi:10.1091/mbc.E06-
11-0990 (2007).
68 Fan, L., Pellegrin, S., Scott, A. & Mellor, H. The small GTPase Rif is an alternative
trigger for the formation of actin stress fibers in epithelial cells. J Cell Sci 123, 1247-
1252, doi:10.1242/jcs.061754 (2010).
69 Pellegrin, S. & Mellor, H. The Rho family GTPase Rif induces filopodia through mDia2.
Curr Biol 15, 129-133, doi:10.1016/j.cub.2005.01.011 (2005).
70 Berg, J. S. & Cheney, R. E. Myosin-X is an unconventional myosin that undergoes
intrafilopodial motility. Nat Cell Biol 4, 246-250, doi:10.1038/ncb762 (2002).
71 Berg, J. S., Derfler, B. H., Pennisi, C. M., Corey, D. P. & Cheney, R. E. Myosin-X, a novel
myosin with pleckstrin homology domains, associates with regions of dynamic actin.
J Cell Sci 113 Pt 19, 3439-3451 (2000).
78 72 Tokuo, H. & Ikebe, M. Myosin X transports Mena/VASP to the tip of filopodia.
Biochem Biophys Res Commun 319, 214-220, doi:10.1016/j.bbrc.2004.04.167
(2004).
73 Tatavarty, V., Kim, E. J., Rodionov, V. & Yu, J. Investigating sub-spine actin dynamics
in rat hippocampal neurons with super-resolution optical imaging. PLoS One 4,
e7724, doi:10.1371/journal.pone.0007724 (2009).
74 Lukyanov, K. A., Chudakov, D. M., Lukyanov, S. & Verkhusha, V. V. Innovation:
Photoactivatable fluorescent proteins. Nat Rev Mol Cell Biol 6, 885-891,
doi:10.1038/nrm1741 (2005).
75 Nienhaus, G. U. et al. Photoconvertible fluorescent protein EosFP: biophysical
properties and cell biology applications. Photochem Photobiol 82, 351-358,
doi:10.1562/2005-05-19-RA-533 (2006).
76 Saxton, M. J. & Jacobson, K. Single-particle tracking: applications to membrane
dynamics. Annu Rev Biophys Biomol Struct 26, 373-399,
doi:10.1146/annurev.biophys.26.1.373 (1997).
77 Zhao, L. P. et al. Cloning and characterization of myr 6, an unconventional myosin of
the dilute/myosin-V family. Proc Natl Acad Sci U S A 93, 10826-10831 (1996).
78 Roland, J. T., Lapierre, L. A. & Goldenring, J. R. Alternative splicing in class V myosins
determines association with Rab10. J Biol Chem 284, 1213-1223,
doi:10.1074/jbc.M805957200 (2009).
79 Hales, C. M., Vaerman, J. P. & Goldenring, J. R. Rab11 family interacting protein 2
associates with Myosin Vb and regulates plasma membrane recycling. J Biol Chem
277, 50415-50421, doi:10.1074/jbc.M209270200 (2002).
79 80 Lindsay, A. J. & McCaffrey, M. W. Rab11-FIP2 functions in transferrin recycling and
associates with endosomal membranes via its COOH-terminal domain. J Biol Chem
277, 27193-27199, doi:10.1074/jbc.M200757200 (2002).
81 Swiatecka-Urban, A. et al. Myosin Vb is required for trafficking of the cystic fibrosis
transmembrane conductance regulator in Rab11a-specific apical recycling
endosomes in polarized human airway epithelial cells. J Biol Chem 282, 23725-
23736, doi:10.1074/jbc.M608531200 (2007).
82 Chu, B. B. et al. Requirement of myosin Vb.Rab11a.Rab11-FIP2 complex in
cholesterol-regulated translocation of NPC1L1 to the cell surface. J Biol Chem 284,
22481-22490, doi:10.1074/jbc.M109.034355 (2009).
83 Fan, G. H., Lapierre, L. A., Goldenring, J. R., Sai, J. & Richmond, A. Rab11-family
interacting protein 2 and myosin Vb are required for CXCR2 recycling and receptor-
mediated chemotaxis. Mol Biol Cell 15, 2456-2469, doi:10.1091/mbc.E03-09-0706
(2004).
84 Provance, D. W., Jr. et al. Myosin-Vb functions as a dynamic tether for peripheral
endocytic compartments during transferrin trafficking. BMC Cell Biol 9, 44,
doi:10.1186/1471-2121-9-44 (2008).
85 Schuh, M. An actin-dependent mechanism for long-range vesicle transport. Nat Cell
Biol 13, 1431-1436, doi:10.1038/ncb2353 (2011).
86 Lise, M. F. et al. Involvement of myosin Vb in glutamate receptor trafficking. J Biol
Chem 281, 3669-3678, doi:10.1074/jbc.M511725200 (2006).
87 Woolner, S. & Bement, W. M. Unconventional myosins acting unconventionally.
Trends Cell Biol 19, 245-252, doi:10.1016/j.tcb.2009.03.003 (2009).
80 88 Lapierre, L. A. et al. Myosin vb is associated with plasma membrane recycling
systems. Mol Biol Cell 12, 1843-1857 (2001).
89 Nelson, S. R., Ali, M. Y., Trybus, K. M. & Warshaw, D. M. Random walk of processive,
quantum dot-labeled myosin Va molecules within the actin cortex of COS-7 cells.
Biophys J 97, 509-518, doi:10.1016/j.bpj.2009.04.052 (2009).
90 Pierobon, P. et al. Velocity, processivity, and individual steps of single myosin V
molecules in live cells. Biophys J 96, 4268-4275, doi:10.1016/j.bpj.2009.02.045
(2009).
91 Niu, L. & Yu, J. Investigating intracellular dynamics of FtsZ cytoskeleton with
photoactivation single-molecule tracking. Biophys J 95, 2009-2016,
doi:10.1529/biophysj.108.128751 (2008).
92 Manley, S. et al. High-density mapping of single-molecule trajectories with
photoactivated localization microscopy. Nat Methods 5, 155-157,
doi:10.1038/nmeth.1176 (2008).
93 Lagerholm, B. C., Weinreb, G. E., Jacobson, K. & Thompson, N. L. Detecting
microdomains in intact cell membranes. Annu Rev Phys Chem 56, 309-336,
doi:10.1146/annurev.physchem.56.092503.141211 (2005).
94 Kerber, M. L. & Cheney, R. E. Myosin-X: a MyTH-FERM myosin at the tips of filopodia.
J Cell Sci 124, 3733-3741, doi:10.1242/jcs.023549 (2011).
95 Schneider, M. E. et al. A new compartment at stereocilia tips defined by spatial and
temporal patterns of myosin IIIa expression. J Neurosci 26, 10243-10252,
doi:10.1523/JNEUROSCI.2812-06.2006 (2006).
81 96 Belyantseva, I. A. et al. Myosin-XVa is required for tip localization of whirlin and
differential elongation of hair-cell stereocilia. Nat Cell Biol 7, 148-156,
doi:10.1038/ncb1219 (2005).
97 Salles, F. T. et al. Myosin IIIa boosts elongation of stereocilia by transporting espin 1
to the plus ends of actin filaments. Nat Cell Biol 11, 443-450, doi:10.1038/ncb1851
(2009).
98 Bear, J. E. & Gertler, F. B. Ena/VASP: towards resolving a pointed controversy at the
barbed end. J Cell Sci 122, 1947-1953, doi:10.1242/jcs.038125 (2009).
99 Schaub, S., Meister, J. J. & Verkhovsky, A. B. Analysis of actin filament network
organization in lamellipodia by comparing experimental and simulated images. J Cell
Sci 120, 1491-1500, doi:10.1242/jcs.03379 (2007).
100 Arrio-Dupont, M., Foucault, G., Vacher, M., Devaux, P. F. & Cribier, S. Translational
diffusion of globular proteins in the cytoplasm of cultured muscle cells. Biophys J 78,
901-907, doi:10.1016/S0006-3495(00)76647-1 (2000).
82