Microtubules provide a structural framework and orchestrate intracellular trafficking

Get Complete Project Material File(s) Now! »

Intermediate filaments maintain cellular integrity

Despite the great diversity of intermediate filament proteins, our knowledge of inter-mediate filament functions is still very young and fragmented. Unlike microtubules and actin filaments, the expression of intermediate filament proteins is extremely dependent on the cell type. Also, intermediate filaments can be found in the cyto-plasm or in the nucleus (the nuclear intermediate filaments are called lamins). In this section, we will only mention cytoplasmic intermediate filaments:
• Because of their unique physical properties, intermediate filaments control cell shape and prevent cells from irreversible mechanical damage. Intermediate filaments are the only cytoskeletal subcomponent which is not polar: they are symmetric and no molecular motor associated with any intermediate filament type is known so far. Also, as indicated in table 1.1, typical persistence lengths of intermediate filaments range from 100 nm to 1 µm. Their persistence length being much smaller than microtubule and actin persistence lengths, they are the most flexible cytoskeletal subcomponent. As a consequence, intermediate filaments play a peculiar role in cell mechanics by providing strength and resistance to cells. For instance, in keratinocytes, keratin loss leads to significant cell softening [Ramms et al. 2013]. Intermedi-ate filaments also seem to exhibit an adapted response to the applied forces. Strain-stiffening of individual vimentin filaments has been measured in vitro: at low forces, intermediate filaments are fully elastic whereas at high forces, they stiffen and deform in a plastic manner [Block et al. 2015]. Intermediate filaments can be seen as a guardian of cellular integrity, providing mechanical resistance to avoid large and fast cellular deformations.
• Intermediate filaments are associated to several linkers and mem-brane-associated proteins and play a crucial role in intracellular or-ganization. Along with microtubules and actin filaments, cytoplasmic inter-mediate filaments are connected to the nucleus by the Linker of Nucleoskele-ton and Cytoskeleton (LINC) complex, an assembly of proteins connected to the nuclear envelope [Dupin & Etienne-Manneville 2011]. It has been shown that intermediate filaments regulate the size, shape and rigidity of the nu-cleus, but also chromatin organization [Keeling et al. 2017]. As previously stated, intermediate filaments are not associated to any molecular motor. Unlike microtubules, they are not major players in vesicular transport. How-ever, intermediate filaments interact with virtually all intracellular organelles, among which mitochondria or the Golgi apparatus, through linking proteins. Keratins, desmin and neurofilaments control mitochondrial location and func-tion, and Golgi positioning is regulated by keratins, vimentin and neurofila-ments, and endosomal/lysosomal protein distribution by vimentin [Toivola et al. 2005].
• Intermediate filaments are involved in many dynamic cellular pro-cesses. Most of our knowledge related to intermediate filaments comes from diseases in which cells have major disregulations of some intermediate filament proteins. Not only are intermediate filament proteins drastically different from one cell type to another, but they also evolve during the cell cycle and dynamic processes within a given cell type. Intermediate filaments has been found to play a crucial role in the regulation of apoptotic events by protecting cells from apoptosis [Omary et al. 2004]. Mutations on some keratins correlate with increased cell apoptosis and liver injury [Omary et al. 2009]. Intermediate fil-aments are also critical in cell migration: in astrocytes, intermediate filaments depletion slows cell migration [Dupin & Etienne-Manneville 2011]. In many cell types, intermediate filaments facilitate cell motility and migration, so that increased intermediate filament protein expression can be associated with can-cer progression and metastasis in some cases [Etienne-Manneville 2018]. Fi-nally, intermediate filaments impact cell size, cell growth and proliferation. Figure 1.4 depicts the role of intermediate filaments in cellular responses to extracellular stimuli.
Figure 1.4: Intermediate filaments: role in cellular responses to extracellular stimuli. From [Etienne-Manneville 2018].

Functions of the cytoskeleton in eukaryotic cells 

Microtubules provide a structural framework and orches-trate intracellular trafficking

Microtubules have become a resarch topic per se during the 1960s. Their association with several molecular motors and their dynamics make microtubules a key player in cell polarity, cell organization and intracellular transport. Here, we give three of the main functions of microtubules:
• Microtubules establish the global polarity of the cell. One of the most important properties of the microtubule network is that they are highly dynamic, constantly polymerizing and depolymerizing, as we will describe in 1.4.1. Microtubule half-time vary during the cell cycle, but it ranges from 10 s to 10 min. Interestingly, cells lacking a proper dynamic microtubule network- e.g. after using drugs affecting its dynamics – are often much less motile and poorly polarized. While actin structures are crucial in cell motility, it has been shown that microtubules induce cortical polarity and regulate actin dynamics [Siegrist & Doe 2007]. Microtubules, rather than establishing cell polarity, ensure its maintenance by mediating the distribution of inhibitory signals [Zhang et al. 2014]. Microtubules also impact the organization of the nucleus and of other organelles. For instance, in Xenopus, microtubule dynamics is balanced spatially and temporally for nuclear formation and its perturbation changes nuclear morphology [Xue et al. 2013].
• Molecular motors associate with microtubules to perform intracel-lular trafficking. As myosins bind actin filaments to transport vesicles along them, two types of molecular motors bind microtubules : kinesins and dynein. Discovered in 1985, kinesins are anterograde transport motors, which means that they transport vesicles, organelles and other cargoes towards the (+) ends of microtubules from the center of a cell to its periphery [Vale et al. 1985]. Fourteen families of human kinesins exist and all share a common amino acid sequence of the motor domain. Conversely, cytoplasmic dynein moves on microtubules towards their (-) ends: it is a retrograde transport motor. Dynein-induced retrograde transport is important to send endocytosis prod-ucts to the center of the cell. As myosins, both kinesins and dynein convert the chemical energy stored in ATP into mechanical work.
• Microtubules spatially regulate mitosis and cytokinesis. We already mentioned the crucial role of actin, making contractile rings, to perform cy-tokinesis. This contraction occurs at the end of mitosis, whereas the first steps of mitosis (and also cytokinesis) are regulated by microtubules. First, micro-tubules determine the cleavage plane and position the site of division. Second, they transport vesicles to the cleavage plane. Microtubules form the mitotic apparatus (composed of the mitotic spindle and of the aster) which pulls the sister chromatides in opposite directions. The construction of this machin-ery is spatially regulated by microtubules but also requires the coordinated activities of many proteins [Straight & Field 2000].

Intermediate filaments

Diversity of intermediate filaments in eukaryotic cells

With more than seventy different proteins, intermediate filaments show a high vari-ability. They are subcategorized in six different types that we will brielfy describe below and which are recapitulated in table 1.2.
Keratins are part of type I (acidic keratins) and type II (neutral or basic keratins). There are more than fifty different keratins in animal cells. They are among the smallest intermediate filament proteins with a molecular weight between 40 kDa and 70 kDa. Keratins are mostly expressed in epithelial cells, in which one type I keratin copolimerizes with one type II keratin to give rise to a keratin filament. Keratins are widespread in the cytoplasm of epithelial cells, but also form specific structures like hair, nails and horns.
Type III intermediate filament proteins include vimentin, which can be found in a lot of cell types, from fibroblasts to endothelial cells. Another well-characterized type III protein is desmin, which can be found in muscle cells. Glial Fibrillary Acidic Protein (GFAP) is specifically expressed in glial cells. All type III interme-diate filament proteins have molecular weights around 55 kDa.
Neurofilaments constitute the type IV intermediate filament proteins. As their name suggests, they are found in neurons, specially in the axons of motor neurons. They are thought to be crucial for these long and thin processes than can reach 1 m to join neurons between them. Type IV intermediate filament proteins are the largest ones with molecular weights up to 240 kDa3.
Notably, type V intermediate filament proteins are not part of the cytoskeleton: they are part of the nuclear envelope and are called lamins. Lamins have a molec-ular weight around 70 kDa
Finally, type VI4 intermediate filament proteins are structurally quite different from the other types, being characterized by a long C-terminal tail. Among them, some proteins are specifically found in lens.
Most of our knowledge regarding intermediate filaments comes from studies of in-termediate filament-associated diseases. For instance, early-onset megalencephaly, progressive spasticity and dementia characterize Alexander disease, which is caused by a mutation on the gene coding for GFAP [Omary et al. 2004]. Mutated desmin cause myopathies, and Charcot-Marie-Tooth disease (characterized by symmetrical muscle weakness, wasting, foot deformitites, difficulty walking, reduced tendon re-flexes) is due to mutations in lamins [Omary et al. 2004]. The research carried out on these diseases have led to greatly improve our understanding of the functions and the diversity of intermediate filament proteins.
3Some type VI intermediate filament proteins – like nestin and synemin – are considered by some papers to be part of type IV because their genomic structure is similar to the neurofilament family and -internexin [Liem 2013].
• The tripartite monomer structure of intermediate filament proteins. Like all proteins, intermediate filament proteins have two ends: an amine group, called the N-terminus and a carboxylic group, called the C-terminus. As they are translated from messenger Ribonucleic Acid (RNA), they are cre-ated from N-terminus to C-terminus. The domain next to the N-terminus of intermediate filament proteins is called the head and the one next to the C-terminus is called the tail. Between these two domains, we can find several -helical structures: they form the rod (see figure 1.5.A). All intermediate filaments share this tripartite monomer structure. Under stress, the rod do-main can stretch and some parts of the -helical domain can be uncoiled and form sheets instead [Qin et al. 2009]. Nuclear intermediate filaments – lamins – also have this structure. Their structural singularity relies on the much longer C-terminal domain in which can be found a Nuclear Localization Signal (NLS), required for transport into the nucleus (see figure 1.5.B-C that compares human lamin and human vimentin structures).
• Assembly of intermediate filament proteins depends on their type. All intermediate filaments are assembled from fibrous proteins that exhibit a central -helical rod domain which facilitates the formation of dimeric coiled-coil complexes. However, their assembly vary a lot depending on the type of intermediate filament protein involved. The first step common to all interme-A. Intermediate filaments are composed of head, rod and tail domains. Adapted from [Lopez et al. 2016] B & C. Structural models of human vimentin, a cytoplasmic intermediate filament protein (B) and human lamin A, a nuclear intermediate filament protein (C). Scale bar: 5 nm. From [Herrmann et al. 2007] D & E. Electron microscopy of recombinant human lamin A (D) and recombinant vimentin (E). Scale bar: 200 nm. Arrow heads indicate prominent « beadings » of the filaments, which are typical structures found when lamins are reconstituted in vitro. From [Herrmann & Aebi 2016].

Intermediate filaments 

diate filament proteins is the formation of a dimer after two monomers bind to each other by their rod domain. The two monomers involved can be identical and form homodimers or they can be different and form heterodimers. Ker-atins, for instance, always form heterodimers by associating a type I keratin protein (acidic) with a type II keratin protein (neutral-basic). The next step varies:
– for cytoplasmic intermediate filament proteins, two dimers form a tetramer in a half-staggered manner through the rod domains aligned in an an-tiparallel orientation. This tetramer is non polar and is called protofila-ment. Then, eight tetramers are aligned laterally to form a unit-length filament. Several unit-length filaments will finally anneal end-to-end to form a non-polar and mature intermediate filament. This process is de-picted in figure 1.6.A.
– regarding lamins, dimers bind their N-terminus to the C-terminus of an-other dimer by a peptide bond, forming a long polar structure. Two of these structures will then align laterally in an anti-parallel orientation to form a protofilament. Protofilaments then assemble to create a symmet-ric filament, which now is also non-polar, like cytoplasmic intermediate filaments (see figure 1.6.B.)
Figure 1.5.D-E shows electron microscopy images of these intermediate fila-ments.
Unlike microtubules and actin filaments, ATP and Guanosine Triphosphate (GTP) are not directly required in the assembly of intermediate filaments because the different steps (monomer to dimer, dimer to tetramer, etc.) do not rely on ATP-bound or GTP-bound monomers. However, immature inter-mediate filaments are transported along microtubules, or actin filaments for most keratins, and therefore require chemical energy for their transport.

READ  La modélisation du séchage et du transfert de gaz

Vimentin, an intermediate filament protein present in many cells

With keratins, vimentin is among the most studied intermediate filament proteins. It can be found in many different cell types, specially in mesenchymal cells. Human vimentin is a 57 kDa protein the structure of which appears in figure 1.5.A-B,E. Here, we review some of the key aspects of vimentin5.
• Vimentin is crucial in the epithelial-mesenchymal transition. Animal tissues can be divided into four different families. Among them, epithelium corresponds to the cells lining the outer surfaces of organs and blood vessels. Epithelial cells are highly polarized and are strongly connected by several junctions between them. Epithelial cells can leave their original tissue and, A : Assembly of cytoplasmic intermediate filaments. Adapted from [Hohmann & De-hghani 2019]. B : Assembly of nuclear intermediate filaments (lamins). Adapted from [Dittmer & Misteli 2011].
after losing their polarity and breaking cell-cell adhesions, become multipo-tent cells called mesenchymal cells. This process is called the Epithelial-Mesenchymal Transition (EMT) and is required for numerous developmental processes, wound healing but also to initiate metastasis during cancer. Gene expression is modified during EMT and, while vimentin expression is quite low in epithelium, it is upregulated in mesenchymal cells. For instance, it has been shown that vimentin promotes EMT phenotypes in breast cancer cells by mediating the expression of slug, an EMT protein [Liu et al. 2015]. Vimentin is not only involved in signalling pathways, but also in migration. Non-metastatic breast cancer cells MCF-7 – an epithelial cell line than nor-mally does not express vimentin – rapidly adopt mesenchymal shapes after vimentin transfection. Conversely, silencing vimentin causes mesenchymal cells to adopt epithelial shapes [Mendez et al. 2010]. Along with these shape transitions, increase in cell motility and in focal adhesions dynamics were mea-sured to be coincident with vimentin filament assembly [Mendez et al. 2010]. In the context of the EMT, vimentin appears to play a major role in the gain of migratory properties, as shown in figure 1.7.A.
• Vimentin in organelle positioning. In addition to increasing migra-tion properties, vimentin filaments also regulate the positioning of several organelles in eukaryotic cells. For instance, vimentin interacts a lot with mitochondria. Cells lacking an intact vimentin filament network exhibit an increased level in the motility of mitochondria [Nekrasova et al. 2011]. The authors suggest than vimentin intermediate filaments bind to mitochondria an anchor them within the cytoplasm. Vimentin has also been shown to interact with the Golgi apparatus [Gao & Sztul 2001] and with melanosomes, forming an intricate cage around them [Chang et al. 2009]. Finally, vimentin has been recently described to form ball-like structures and rings around the nucleus during the first step of adhesion that are able to strongly deform the cell nu-cleus [Terriac et al. 2019]. These features are schematically recapitulated in figure 1.7.B.
• Vimentin in astrocytes and glioma cells. As my PhD resorts to a cell line derived from a human malignant astrocytoma6, we focus here on vimentin – as well as its links to other intermediate filament proteins – in astrocytes and glioma cells. Like in other cancers [Satelli & Li 2011], vimentin is upregulated in glioma and glioblastoma, a grade IV glioma. Interestingly, using withaferin-A (a chemical inhibitor of vimentin) in glioblastoma induces glioblastoma cell morphology changes, inhibits the motility of glioblastoma cells and leads to a reduction of glioblastoma cell growth in vitro [Zhao et al. 2018]. More than being a marker of tumour in glial cells, vimentin is a marker of a poor outcome in gliobastoma patients [Zhao et al. 2018].
Another type III intermediate filament protein is expressed in astrocytes (and is even specific for glial cells): the Glial Fibrillary Acidic Protein (GFAP). Even though studies on the link between GFAP expression and malignant phe-notypes or tumour growth are controversial, it has been shown long ago that 6 The brain is mainly made of two cell types – neuronal cells and glial cells. Among glial cells, we can find several subtypes, such as astrocytes, oligodendrocytes, etc.. Tumours derived from these cells are called glioma, astrocytoma, oligodendrocytoma, etc.
A: Epifluorescence image of nestin, GFAP and vimentin. Scale bars: 20 µm. B: 3D structured illumination microscopy picture. Scale bars: 10 µm (main image) and 1 µm (inset). Both fluorescence intensity profiles were obtained along the corresponding dotted arrow. From [Leduc & Etienne-Manneville 2017].
human gliomas co-express GFAP and vimentin [Herpers et al. 1986] and that both proteins copolimerize in the same intermediate filament system [Wang et al. 1984]. More recently, epifluorescence microscopy and 3D structured illu-mination microscopy images have shed light on the distribution of cytoplasmic intermediate filaments in astrocytes [Leduc & Etienne-Manneville 2017]. In addition to GFAP and vimentin, the authors studied another intermediate filament protein: nestin. Figure 1.8.A shows that astrocytes have a heteroge-neous expression of intermediate filament proteins but their distributions are similar. Besides, they strongly colocalize in cells expressing several intermedi-ate filament proteins. Figure 1.8.B uses a superresolution imaging technique to show that single filaments are composed of several intermediate filament proteins.



Dynamics of microtubules

Unlike intermediate filaments, microtubules are extremely dynamic: they exhibit a fast turnover in living cells. In 1986, in vitro experiments showed that 80% of the microtubules in interphase cells turn over in 15 min [Schulze & Kirschner 1986]. Microtubules exhibit a cyclic behaviour, oscillating between long and progressive phases of polymerization and short and brutal depolymerization events. This pro-cess has been named dynamic instability and will be defined below.
• Microtubule polymerization is initiated by nucleation on pre-existing seeds7. If recruiting heterodimers of -tubulin to elongate an already formed microtubule is an energetically favourable process which explains why microtubules polymerize fast, initiating microtubule polymerization is a highly unfavourable process. Hence, initiation is the rate-limiting step in microtubule polymerization. In vitro, microtubule initial growth progresses slowly, as it proceeds from small entities for which dissassembly is energetically favoured over assembly (see figure 1.9.A). In cells, some preformed nuclei are found at Microtubule-Organizing Centers (MTOCs). For instance, -tubulin ring complexes ( -TuRC) are made of -tubulin, another member of the tubulin family. These complexes are the structural basis of the thirteen protofilament structure of microtubules (see figure 1.9.B). In bulk assembly assays, the pres-ence of preformed nuclei increases the fraction of polymerized microtubules with time (see figure 1.9.C) [Kollman et al. 2011].
• Growing and shrinking, the dynamic instability of microtubules. The complexity of microtubule dynamics goes further than nucleation-elonga-tion processes. While microtubules elongate, their polymerization is regularly and abruptly stopped by depolymerization phases. The transition point be-tween polymerization and depolymerization is called catastrophe. Conversely, when a microtubule is (quickly) depolymerizing, it can switch back to a poly-merization phase by an event called rescue (see figure 1.9.D-E). Microtubules are therefore constantly growing and shrinking [Mitchison & Kirschner 1984], and what governs the transitions between these phases – especially rescues – is still not totally established.
Catastrophes have long been viewed as the result of the loss of a protective end structure [Mitchison & Kirschner 1984] [Desai & Mitchison 1997]. Un-der this hypothesis, the probability of a catastrophe event has to be constant over time. However, this does not match the measurements. It is now ac-cepted that catastrophes are not a single step process but a multiple step process, which implies that the probability of undergoing a catastrophe in-7 In many studies, this nucleation seed is simply called nucleus. We have used here the word seed- as in thermodynamics – to introduce the concept and avoid any confusion with the cell nucleus. In the following, we use the word nucleus to better match the literature.
A: De novo formation of microtubules from heterodimers of αβ-tubulin is energetically un-favourable until a sufficiently large oligomer is formed (after N steps). B: Preformed nuclei (such as the -TuRC complex) allow microtubule to bypass the slow phase in vivo. C: Influence of the presence of a preformed nucleus on the polymerization kinetics. Adapted from [Kollman et al. 2011]. D: Microtubule length against time depicting dynamic instability and showing catastrophes and rescues. Adapted from [Mauro et al. 2019]. E: Schematic representation of dynamic instability of microtubules. (1) Closure of the terminal sheet structure generates a metastable, blunt-ended microtubule (2) which may pause, undergo further growth or switch to the depolymerization phase. A shrinking microtubule is characterized by fountain-like arrays of ring and spiral protofilament structures (3). This cycle is completed by exchanging GDP of the disassembly products with GTP (4). Adapted from [Dráber et al. 2012].
creases with time: microtubule catastrophe can be viewed as an aging pro-cess [Gardner et al. 2013]. The rescue process remains much more debated. It was shown in vitro that rescue events increase with free tubulin concen-tration [Walker et al. 1988]. More recently, after observing with a specific antibody that microtubules host some GTP-tubulin islands (or remnants) within their lattice, it has been hypothesized that rescue events initiate from these remnants [Dimitrov et al. 2008]. Molecular dynamics simulations and in vitro experiments show that GTP-tubulin remnants regulate the kinetics of depolymerization [Bollinger et al. 2020]. But how these remnants appear in the lattice is still an open question.

Table of contents :

1 The cytoskeleton 
1.1 A network of crosslinked filamentous biopolymers
1.2 Functions of the cytoskeleton in eukaryotic cells
1.2.1 Actin filaments regulate cell shape and movement
1.2.2 Intermediate filaments maintain cellular integrity
1.2.3 Microtubules provide a structural framework and orchestrate intracellular trafficking
1.3 Intermediate filaments
1.3.1 Diversity of intermediate filaments in eukaryotic cells
1.3.2 Structure and assembly of intermediate filaments
1.3.3 Vimentin, an intermediate filament protein present in many cells
1.4 Microtubules
1.4.1 Dynamics of microtubules
1.4.2 Microtubules and post-translational modifications
2 Mechanical properties of the cytoskeleton 
2.1 Measuring mechanical properties in cell biology
2.1.1 Cytoskeletal mechanics measurements in vitro
2.1.2 Whole-cell-scale, cortical and intracellular force measurements
2.1.3 Our approach: combining optical tweezers-based intracellular rheology with live cell imaging
2.2 Mechanics of microtubules in vitro
2.2.1 Anisotropic stiffness of microtubules
2.2.2 Variability in the measurements of microtubule flexural rigidity
2.2.3 Microtubule response to repeated mechanical stress
2.3 Mechanics of intermediate filaments in vitro
2.3.1 Networks of intermediate filaments: highly deformable and almost unbreakable
2.3.2 Individual intermediate filaments exhibit nonlinear strain-stiffening
2.4 Microtubules and intermediate filaments in cellulo
2.4.1 Measuring mechanics of cytoskeletal filaments in cellulo
2.4.2 Mechanical contribution of microtubules in cells
2.4.3 Mechanical contribution of intermediate filaments in cells
3 Mechanical coupling within the cytoskeleton 
3.1 Cytoskeletal crosstalk involving vimentin intermediate filaments and microtubules
3.1.1 Crosstalk through molecular motors
3.1.2 Crosstalk through crosslinking proteins
3.2 Impact of cytoskeletal crosstalk on network organization and cell mechanics
3.2.1 Synergistic organization of cytoskeletal networks
3.2.2 Mechanical reinforcement mediated by cytoskeletal interactions
4 Aims of the PhD project
5 Materials and Methods 
5.1 Cell culture
5.2 Immunofluorescence staining and fixed cell imaging
5.3 Live cell imaging of vimentin and tubulin in cellulo
5.4 Drugs targeting microtubules
5.5 ATP depletion
5.6 Optical tweezer-based microrheology
5.7 Data analysis: from the movies to the effective stiffness
5.8 Statistical tests
6 Results and Discussion
6.1 Mechanics of vimentin bundles and microtubules
6.1.1 In cellulo, vimentin bundles are stiffer than microtubules
6.1.2 Sequential deflections make vimentin bundles more rigid
6.2 Mechanical coupling between microtubules and vimentin intermediate filaments
6.2.1 The vimentin network does not play a key role in the mechanical properties of microtubules
6.2.2 Modifying microtubule stability affects vimentin mechanical behaviour
6.3 Study of a post-translational modification: acetylation
6.3.1 Acetylation leads to microtubule softening
6.3.2 Acetylated microtubules impact vimentin bundle mechanics
6.4 Preliminary results: role of ATP in cytoskeletal mechanics
7 Conclusion and Perspectives


Related Posts