Movie plays in real time. The field of view is 20 µm in the horizontal direction. Stationary avidin-coated magnetic beads (extreme right) label the minus end of the microtubule. The bead is 500 nm in diameter.
Movie plays in real time. The field of view is 25 µm in the horizontal direction. Stationary avidin-coated magnetic beads (extreme right) label the minus end of the microtubule. A large vesicle moves over a long distance toward the microtubule minus end. Another smaller vesicle is also seen to move toward the plus end of the same microtubule. The vesicles appear to cross each other on the same microtubule, presumably because the motors move on adjacent protofilaments of the MT. A faint stationary spot at the center of the field of view is from an optical trap. Both vesicles walked out of the optical trap because the laser power was low. Several vesicles are observed to diffuse around in the background. These are not attached to the microtubule.
Transport of intracellular organelles along the microtubule cytoskeleton occurs in a bidirectional manner due to opposing activity of microtubule-associated motor proteins of the kinesin and dynein families. Regulation of this opposing activity and the resultant motion is believed to generate a polarized distribution of many organelles within the cell. The bidirectional motion can be reconstituted on in vitro assembled microtubules using organelles extracted from cells. This provides an opportunity to understand the regulation of intracellular transport through quantitative analysis of the motion of organelles in a controlled environment. Such analysis requires the use of polarity-labeled microtubules to resolve the plus and minus components of bidirectional motion. However, existing methods of in vitro microtubule polarity labeling are unsuitable for high-resolution recording of motion. Here we present a simple and reliable method that uses avidin-coated magnetic beads to prepare microtubules labeled at the minus end. The microtubule polarity can be identified without any need for fluorescence excitation. We demonstrate video-rate high-resolution imaging of single cellular organelles moving along plus and minus directions on labeled microtubules. Quantitative analysis of this motion indicates that these organelles are likely to be driven by multiple dynein motors in vivo.
Long-distance intracellular transport relies on the microtubule (MT) cytoskeleton and MT-associated molecular motor proteins of kinesin and dynein families. MTs are polymerized protein filaments of α and β tubulin with a fast-growing plus end extending toward the cell periphery, and a minus end juxtaposed to the cell nucleus in interphase cells (1). MT-associated motor proteins are mechanochemical enzymes that carry cellular organelles (e.g., mitochondria, endosomes, melanosomes) as cargo along MTs using energy derived from ATP hydrolysis (1,2,3). Motor proteins of the kinesin and dynein families usually move toward MT-plus and minus directions within the cell respectively. Such MT-based motion is often bidirectional because of the presence and simultaneous activity of both kinds of motors on an individual cargo (1,4,5). The motors are hypothesized to exist on individual cargos within an “intracellular transport complex” that also includes non-motor proteins such as dynactin (6). This complex is believed to regulate net motion of the cargo in a manner that is poorly understood, but a topic of intense debate (1,4,5,7,8). The regulation of motion is important to understand, because it is largely responsible for the distribution of a multitude of organelles to distinct cellular locations. Live imaging and quantitative analysis of organelle motion inside a cell presents considerable difficulty, and has been demonstrated only in a few systems (7,9,10). An attractive alternative is to reconstitute the bidirectional motion of cellular organelles in controlled in vitro assays. Such reconstitution assays have been demonstrated for squid organelles (11), herpes virus (12), endosomes (13,14), vesicles (15,16), melanosomes (17), and other MT-motor driven cargos (18). If these assays are combined with high-resolution motion analysis, there is an excellent opportunity to understand the role of specific motor and non-motor proteins in the regulation of bidirectional intracellular transport.
To date, it has only been possible to obtain high-resolution data of in vitro motor-driven motion from experiments using motor-coated plastic beads. Analysis of the motion of beads has revealed biophysical properties of single motors such as processivity, velocity, step size, and response to load (19,20,21). Since these experiments use a purified motor, motion along the MT is unidirectional and predetermined (toward plus or minus end). It is therefore not necessary to have prior knowledge of the polarity of an MT on which motion is being assayed. In contrast to this, an organelle extracted from the cell can move in a bidirectional manner on the MT due to activity of both plus-and minus-directed motors. To identify the plus and minus components of motion and interpret it in terms of the activity of kinesin and dynein motors respectively, a priori knowledge of MT polarity is required.
Presently, two methods exist for the in vitro labeling of MT polarity. The first method relies on nucleation and extension of MTs from axonemes/centrosomes (16), which is then identified as the minus end of an MT. The use of this method has been limited to a few groups because the isolation of functional axonemes/centrosomes is difficult. In addition, MTs assembled from centrosomes are found to be bundled and overlapping, making it difficult to ascertain direction and follow the motion of organelles over a long distance on a single MT (22). A second method uses fluorescently labeled tubulin to make a brightly fluorescent MT seed (23). The seed is then preferentially extended toward the MT plus end by adding a mixture of fluorescent and non-fluorescent tubulin. This results in a dimly fluorescent MT with a bright minus-end segment, and has been the popular method for microtubule polarity labeling for almost two decades. Unfortunately, this method is not suitable for high-resolution biophysical characterization of motor motion because of the following reasons: (i) simultaneous observation of dimly fluorescent MTs and non-fluorescent organelles in the same field is difficult, (ii) MT fluorescence gets quenched over the period of the assay, (iii) a long exposure (usually using an expensive camera) is required to image the weakly fluorescent MTs (finer details of bidirectional motion such as rapid reversals are therefore lost due to under-sampling), (iv) fluorescently labeled MTs are known to break up into fragments upon exposure to light (24), and (v) intercalation of a fluorescent molecule into the MT could have unknown effects on motor function on the MT.