Project Topics
Get Complete Project Material File(s) Now! »
The architecture of the actin core and actin dynamics in the stereocilia
The actin cytoskeleton is a network of semi-flexible filaments that are active polymers. The filaments can elongate or shrink depending on the surrounding environment. As a result, the network can continuously reorganize to adapt to changing conditions. The actin cytoskeleton is involved in various key cellular processes such as motility, morphogenesis, polarity, transport and cell division (see the reviews of (Laurent Blanchoin et al. 2014) and (Banerjee, Gardel, and Schwarz 2020) regarding the adaptive nature of the actin cytoskeleton in various biological systems). The hair bundle provides a striking example of an actin-based structure that must be tightly regulated during development and maintained at mature stages according to its function as a frequency-selective detector of periodic mechanical stimuli.
Introduction to actin polymerization and its dynamics
The actin protein exists in two forms: the monomeric globular form (G-actin) and the polymerized filamentous form (F-actin). The actin monomer is a 43-kDa globular protein (G-actin) composed of 4 sub-domains that bind to the nucleotides ATP and ADP as well as to Mg2+ ions (Kabsch et al. 1990). G-actin, which has a diameter of ~5-6 nm (Jonathon Howard 2001), is asymmetrical. When individual monomers assemble into a proto-filament, the subunits are orientated in the same direction in a head-to-tail manner. This gives rise to the polarity of the actin filament; the ends of the filament are structurally different. Two parallel F-actin filaments form a double-stranded helix with a 37-nm repeat and a diameter of an equivalent cylinder of about 8 nm. Sub-domains 1 and 3 of a monomer are called the plus (+) or barbed end and sub-domains 2 and 4 constitute the minus (-) or pointed end. The addition of an actin monomer to a filament extends the overall length by 2.7 nm (Jonathon Howard 2001; Phillips 2013) (Fig. I-15.A-C). The assembly of an actin filament induces a conformational change in the actin subunit: G-actin has a twisted conformation whereas F-actin has a flat conformation, obtained by a rotation of 20 degrees of subdomains 1-2 with respect to subdomains 3-4 (Fig. I-15.D-E). The flat conformation is essential for the stability of helical F-actin (Oda et al. 2009).
Actin polymerization kinetics
The polarity of an actin filament is associated with different kinetics of polymerization and depolymerisation at both ends. At saturating concentrations of G-actin (~mM), the barbed end is the fast-growing end, whereas the pointed end is the slow-growing end. Correspondingly, the association rate of ATP-G-actin (k+) is ten times higher at the barbed end than that at the pointed end (T D Pollard 1986). The association and dissociation rates that tune polymerization and depolymerisation at both ends of the actin filament are controlled by the nucleotides (Fig. I-16). In cells, where ATP molecules are abundant, actin filaments are mostly assembled from ATP-G-actin since G-actin has a strong affinity for ATP. Upon incorporation into a filament, an actin subunit hydrolyses its nucleotide at a relatively fast rate of 0.3 s−1 (Laurent Blanchoin and Pollard 2002). ATP-F-actin is then transformed into ADP-Pi (Pi: inorganic Phosphate) F-actin and then into ADP-F-actin. The release of Pi, which is 100 times slower than ATP hydrolysis (Carlier and Pantaloni 1986).
Actin dynamics in stereocilia.
Mammalian hair cells do not regenerate, which means that the hair-bundle must architecture must be maintained throughout the life of the animals. Maintenance and repair of the stereociliary structure result from the incorporation of cytoplasmic G-actin into the paracrystalline actin core. How dynamic is actin in stereocilia? Early studies based on transfection with -actin-GFP of neonatal rata and mouse hair cells in culture had suggested that actin in stereocilia is continuously renewed in 48-72 hr by a treadmilling mechanism, with polymerization at the tip of the stereocilia and depolarization at the base (Schneider et al. 2002; Rzadzinska et al. 2004). Interestingly the treadmilling rate was observed to be proportional to the stereociliary length so that stereocilia of different lengths would be renewed in the same amount of time. However, this attractive model of dynamic maintenance was contested by more recent experiments from several groups and using various approaches in adult frog saccular hair cells and neonatal mice utricular hair cells. Instead, stereocilia were found to be remarkably stable over most of their length, with actin turnover happening only within about 0.5 μm from the stereociliary tips (Zhang et al. 2012; Drummond et al. 2015; Narayanan et al. 2015) (Fig. I-18). To reconcile these contrasting observations, it was proposed that the previously observed propagation of β-actin GFP from the tip to the base of the stereocilia (Schneider et al. 2002; Rzadzinska et al. 2004) might correspond to the incorporation of actin into nascent stereocilia during the development of immature hair bundles or ‘stereociliogenesis’ (Drummond et al. 2015).
Capping and severing proteins
Capping proteins control actin polymerization/depolymerisation by blocking the association or dissociation of actin monomers at the barbed end of actin filaments. Severing proteins generate new barbed ends at break points of each filament, thus increasing the availability of free barbed end for polymerization and depolymerization. Several capping and severing proteins have been identified in the stereocilia. These proteins are not only required for maintaining the proper stereociliary length but also for maintaining the stereociliary width. In stereocilia, capping proteins require myosin motors for transport to the stereociliary tips.
Eps8/Myosin-15/Whirlin – Under normal conditions, MYO15A co-localizes at the stereocilia tips with whirlin (WHRN), its cargo protein (Inna A. Belyantseva et al. 2005; Delprat et al. 2005). Shaker-2 mice, which carry a mutation in the motor domain of MYO15A, have hair bundles that are abnormally short and have lost their staircase pattern. In the whirler mice, for which there is a mutation in the WHRN gene, stereocilia are shorter and wider than those of the control (Mogensen, Rzadzinska, and Steel 2007). In the shaker-2 mice, the capping protein Eps8 fails to localize at the tips of the stereocilia tallest row suggesting that the MYO15A may be responsible for delivering Eps8 onto the barbed end at the stereociliary tip. In Eps8 knockout mice, the hair bundles of both the inner and outer hair cells are abnormally short (Manor et al. 2011) (Fig. I-24). These observations suggest that Eps8/MYO15A/WHRN complex favours the elongation of stereocilia.
Pharmacological perturbation of the mechano-electrical transduction machinery
Blocking the transduction channels – One can block the ion channels that mediate the mechano-electrical transduction using pharmacological drugs, such as benzamil or tubocurarine (Rüsch, Kros, and Richardson 1994; Farris et al. 2004). Channel blocking is demonstrated by measuring the transduction current of a hair cells upon the application of the pharmacological drugs. Stimulating a hair bundle under the control conditions elicits an inward transduction current (up to a few hundreds of picoampere) but in the presence of channel blockers, the transduction current is diminished (Fig. II-3.A). The drugs are thought to work as open channel blockers, meaning that they plug the channels’ pores. Some of these drugs, e.g. benzamil (Fig. II-3.B), may still work as a permeant blockers because the transduction current does not entirely vanish even at high concentration of the drugs.
During my PhD, I incubated for 1 hr the excised sensory tissue in artificial perilymph supplemented with benzamil, a derivative of amiloride, at a concentration of 30 μM or tubocurarine at a concentration of 100 μM) (Sigma Aldrich). For a longer experiment (incubation time 1 – 48 h), I used instead the culture medium described before. The concentration used was based on the dose-response curves to block more than 90% of the MET current (Rüsch, Kros, and Richardson 1994; Farris et al. 2004) (Fig. II-3.B and C). After the incubation, the tissues were rinsed then fixed for electron microscopy or for physiological studies.
Stimulation with flexible fibres and stiffness measurement
The hair-cell bundles can be mechanically stimulated by using a flexible glass fibre of a known stiffness. A flexible fibre is fabricated by pulling a 1.2 mm Ø borosilicate glass capillary (TW120F-3, World Precision Instruments) perpendicular to the axis of the shaft by a micro forge. The fibres are 0.5 – 1 μm in diameter and have a length of 100 – 500 μm. Sputter-coating of the fibres with gold-palladium was done to improve their optical contrast. In an experiment, a flexible fibre is attached to a piezoelectric actuator and submerged under a liquid (artificial perilymph or endolymph). An image of the fibre is then projected onto the centre of a two-quadrant photodiode to measure displacements at a nanometre range (Fig. II-4). The stiffness Kf and drag coefficient ζf of fibre can then be extracted from a spectral analysis of the fibre’s Brownian motions by fitting with a Lorentzian function (Fig. II-5). The stiffness and drag coefficient of fibres used in this study is in the range of 200 – 400 μN·m-1 and 200 – 250 nN·s·m-1.
Table of contents :
ACKNOWLEDGEMENTS
GENERAL INTRODUCTION
I. INTRODUCTION
A. HAIR-CELLS: THE MECHANO-ELECTRICAL TRANSDUCERS
B. THE ARCHITECTURE OF THE ACTIN CORE AND ACTIN DYNAMICS IN THE STEREOCILIA
C. OTHER ACTIN-BASED PROTRUSIONS
D. CONTROL OF STEREOCILIA DIMENSIONS BY MECHANOTRANSDUCTION
II. MATERIALS AND METHODS
A. EXPERIMENTAL PREPARATION OF THE SENSORY TISSUES
B. PHARMACOLOGICAL PERTURBATION OF THE MECHANO-ELECTRICAL TRANSDUCTION MACHINERY
C. MECHANICAL STIMULATION OF HAIR BUNDLES
D. SCANNING ELECTRON MICROSCOPY
E. TRANSMISSION ELECTRON MICROSCOPY
F. IMMUNOLABELLING AND IMMUNOFLUORESCENCE MICROSCOPY
G. INHIBITION OF FORMINS
III. RESULTS
A. MORPHOLOGICAL CHARACTERIZATION OF SACCULAR HAIR BUNDLES OF THE FROG RIVAN
B. EFFECTS OF PHARMACOLOGICAL PERTURBATION OF MECHANO-ELECTRICAL TRANSDUCTION
C. PROBING THE ROLE OF FORMINS FOR ACTIN POLYMERIZATION IN STEREOCILIA
IV. DISCUSSION AND CONCLUSIONS
A. COMPARING RIVAN 92’S TO AMERICAN BULLFROGS’ HAIR CELLS
B. STEREOCILIA WIDENING AND SHORTENING UPON PERTURBATION OF THE MECHANO-ELECTRICAL TRANSDUCTION MACHINERY
V. APPENDIX
A. LATTICE PACKING OF CIRCULAR DISKS
B. SCANNING ELECTRON MICROSCOPY: SAMPLE PREPARATION
C. TRANSMISSION ELECTRON MICROSCOPY
BIBLIOGRAPHY
