Strategies to Identify Tracks and Motors

Historically, experiments with drugs that depolymerize or stabilize actin filaments or microtubules (see Boxes 33-1 and 34-1) provided the first clues about the cytoskeletal polymers that support various biological movements. Identifying the participating motors, if any, has been much more challenging, given a minimum five myosin genes, six kinesin genes, and one dynein gene in budding yeast and about ten times more motor protein genes in humans. A limited number of pharmacologic agents (Box 37-1) can implicate some motors, but the most definitive approach is to alter motor activity or abundance genetically or by RNAi-mediated depletion. Some motors are essential, so deletion mutations are lethal. Conditional mutations have allowed many definitive tests for the functions of these motors in yeast and, in a few cases, in more complex organisms. Other motors are not essential in the sense that cells have alternative strategies that can compensate if the protein is missing or inactive; nevertheless, the protein may have an important function. For example, dynein contributes to mitosis, but Drosophila tissue culture cells that are depleted of dynein can complete mitosis, though only after a long delay during which other motors take over.

Rapid Movements along Microtubules

Organelles in most cells are capable of moving at relatively high velocities, on the order of 1 mms (Table 37-1) along linear microtubule tracks. Thus, the organization of microtubules determines the patterns of these movements (Fig. 37-1; also see Fig. 34-2). Such movements are typically intermittent.

Microtubule-based motor proteins, kinesins and dyneins, power organelle movement along microtubules. Kinesins move organelles toward microtubule plus ends, which are located near the periphery of cells with microtubules radiating from centrosomes. Dyneins are responsible for organelle movements toward the minus ends of microtubules, located at the cell center. Intraflagellar transport of proteins in cilia and flagella (see Fig. 38-20) shares many features with the movements of organelles. Movements of organelles along microtubules have been reconstituted with purified dynein and kinesin. Final proof of the responsible motors usually depends on genetic tests.

An assembly of proteins called the dynactin complex regulates the ability of dynein to transport membranes along microtubules (Fig. 37-2). This complex consists of a short filament of the actin-related protein Arp1 and seven other subunits, including heterodimeric capping protein. The 150-kD subunit (p150^glued) binds to an intermediate chain of dynein. The Arp-1 filament interacts with spectrin associated with the membrane. Mutations in Drosophila p150^glued cause developmental defects in the eye and brain. Some patients with inherited forms of motor neuron degeneration also have mutations of p150^glued.

Figure 37-2 attachment of cytoplasmic dynein to membranes by the dynactin complex. Dynein moves toward the minus end of a microtubule. It is linked to a vesicle by the dynactin complex associated with spectrin on the membrane surface. The dynactin complex consists of a short filament of Arp1 (capped on its barbed end by the capping protein), p50 dynamitin, and by p150^glued, which binds both Arp1 and microtubules.

(Redrawn from Holleran E, Karki S, Holzbaur EL, et al: The role of the dynactin complex in intracellular motility. Int Rev Cytol 182:69–109, 1998.)

Several different proteins link kinesins to transported membranes. Each link is intriguing for reasons independent of intracellular motility. For example, the transmembrane amyloid precursor protein not only binds directly to kinesin-1 light chain but in Alzheimer’s disease also is cleaved to produce amyloid-b peptide—a toxic peptide that is implicated in neuronal death. Another transmembrane protein, JIP-3, binds MAP kinase cascade kinases in addition to kinesin-1 light chains. Mutations in the gene for Drosophila JIP-3 disrupt axonal transport and clog axons with clumps of vesicles. Adapter proteins link kinesin-1 light chains to transmembrane receptor proteins or AMPA glutamate receptors. A PH-domain on kinesin-3 targets the protein to membrane polyphosphoinositides and promotes the formation of dimers, which makes the motor processive. These examples illustrate that many mechanisms exist to link specific kinesin motors to a wide variety of transported vesicles and that much interesting biology will emerge from further characterization.

Fast Axonal Transport

Analysis of microtubule-based movements is particularly favorable in axons of nerve cells because axons are long (up to 1 m) but narrow, the microtubules have a uniform polarity, and organelles move at steady rates in both directions. Furthermore, nerve cells contain high concentrations of microtubules and microtubule motors; indeed, cytoplasmic tubulin, cytoplasmic dynein, and kinesin were all originally isolated from brain.

High-contrast light microscopy of living axons reveals that most membrane-bound organelles move either toward (anterograde) or away from (retrograde) the end of the axon (Fig. 37-3) with some pauses and even occasional changes of direction. Retrograde movements (2.5 mm s⁻¹ or 22 cm/day) are faster than anterograde movements (0.5 mm s⁻¹ or 4 cm/day). At these rates, a round trip from a cell body in the spinal cord of a human to the foot and back takes only three weeks. This might seem slow, but if a 0.1-mm vesicle were the size of a small car, it would be moving anterograde at 50 miles per hour and retrograde at 250 miles per hour. In the axons of vertebrate neurons, mitochondria and autophagic vesicles move back and forth in both directions. Their net movement toward the nerve terminal or cell body depends on physiological conditions.

Figure 37-3 fast transport in cytoplasm isolated from squid giant axons. A, Three frames from a series of video-enhanced differential interference contrast micrographs show movement of organelles in both the anterograde (right) and retrograde (left) directions. Four large organelles are marked with numbers and colored green at zero time, blue at 3 seconds, and red at 5 seconds. Movement (arrows in right panel) is from the white to the black number. The original video record shows hundreds of smaller organelles moving steadily in either an anterograde or a retrograde direction at 1 to 2 mm/second. B, Electron micrograph of a thin section showing vesicles associated with microtubules in axoplasm.

(A, Courtesy of S. Brady, University of Texas Southwestern Medical School, Dallas. B, Courtesy of R. H. Miller, Case Western Reserve Medical School, Cleveland, Ohio.)

Biochemical reconstitution showed that the plus-end motors of the kinesin family are responsible for anterograde movements toward the nerve terminal and the minus-end motor dynein is responsible for movement in the retrograde direction. Proofs of function are more complicated in animals, owing to multiple plus-end motors with partially overlapping functions. Nevertheless, kinesin mutations in flies result in paralysis of the back half of larvae, because transport fails in the longest axons. Mutations in three different kinesin genes also cause human nerve degeneration. Point mutations resulting in amino acid substitutions in the mouse dynein heavy chain cause apparently mild defects in retrograde axonal transport in motor neurons, but the affected neurons die with pathology similar to that of human motor neuron diseases.

Classic nerve ligation experiments revealed the cargo carried in each direction by fast transport (Fig. 37-4). When mechanical constriction blocks transport, diferent organelles pile up on either side. Small, round, and tubular vesicles, including components of synaptic vesicles (see Fig. 11-8), accumulate on the side near the cell body. Fast anterograde transport carries these cargoes from the cell body toward the end of the axon, where they enter the cycle of synaptic vesicle turnover. Endosomes and multivesicular bodies moving by fast retrograde transport pile up on the distal side of the constriction. Retrograde transport can also move signals from nerve terminals to the cell body. For example, when nerve growth factor activates the TrkA receptor tyrosine kinase (see Fig. 24-4) at nerve terminals, the activated receptor is taken up by endocytosis and transported in endosomes to the perinuclear region, where the MAP kinase pathway regulates cell growth. Some viruses also move by retrograde transport.

Figure 37-4 electron micrographs showing the result of nerve ligation. A, The cytoplasm proximal to the ligation demonstrates the accumulation of vesicles and mitochondria, which were being transported toward the nerve terminal to the right. B, The cytoplasm distal to the ligation shows the accumulation of lysosomes, multivesicular bodies, and mitochondria, which were being transported toward the cell body to the left.

(B, Reproduced from Hirokawa N, Sato-Yoshitake R, Yoshida T, Kawashima T: Brain dynein (MAP1C) localizes on both anterogradely and retrogradely transported membranous organelles in vivo. J Cell Biol 111:1027–1037, 1990, by copyright permission of The Rockefeller University Press.)

Dynein piles up on both sides of nerve ligatures. It is associated with vesicles moving in the anterograde direction (toward microtubule plus ends). At nerve terminals, an unknown mechanism activates dynein and reverses the direction of transport. Vesicles that move on microtubules can transfer to and move along actin filaments, using myosin to power local movements at the nerve terminus and in the cortex of the axon.

Many questions remain regarding the control of microtubule motors during fast transport, including how kinesins remain active and dynein remains inactive during the long trip to the nerve terminal, how cytoplasmic dynein on retrograde cargo is activated locally at the nerve terminal and kept active during movement to the cell body, how the bidirectional movements of mitochondria are biased by the physiological state of the cell to achieve net transport, and how defects in fast transport may contribute to neurodegenerative diseases.

Slow Transport of Cytoskeletal Polymers and Associated Proteins in Axons

Many neuronal proteins move slowly from their site of synthesis in the cell body toward the ends of axons and dendrites. This transport is essential, as most protein synthesis occurs in the cell body, whereas more than 99% of cell volume can be in axons and dendrites. If nerve cells were smaller or less asymmetrical, we might not even notice such slow movements. These movements along axons can be followed by labeling the proteins with radioactive amino acids during their synthesis in the cell body (Fig. 37-5A). Proteins that are moved by slow axonal transport are classified into two groups based on their velocities. Tubulin, intermediate filament proteins, and spectrin, which compose the “slow component–a,” move exceedingly slowly, about 0.1 to 1.0 mm per day (or 1 to 10 nm s⁻¹). In a human, these molecules take more than 3 months to travel from their site of synthesis in the spinal cord to the foot. “Slow component–b” moves about 10 times faster and includes 10 times more protein than slow component–a. It is a heterogeneous mixture of proteins, including clathrin, glycolytic enzymes, and actin.

Figure 37-5 experiments on slow axonal transport. A, Pulse-chase experiment. Radioactive amino acids are injected into the spinal cord or eye of an experimental animal. In the nerve cell body, radioactive tracer is incorporated into proteins, which are transported along the axon. Some proteins are incorporated into stationary structures and are left along the way. B, Photobleaching experiment. A cultured nerve cell is injected with tubulin labeled with a fluorescent dye. Tubulin fills the cytoplasm and axon as it grows out. A section of the axon is then bleached with a strong pulse of light. This bleached zone is stationary over a period of minutes. C, Fluorescence micrographs of the axon of a cultured rat neuron showing rapid transport of a neurofilament labeled with subunits fused to GFP. Note the photobleached region (bracket) and the ends of the moving neurofilament (arrows). The neurofilament moves rapidly into the bleached region, but the bleached region does not move because most of the neurofilaments are stationary. Scale bar is 5 mm.

(A–B, Redrawn from Cleveland DW, Hoffman PN: Slow axonal transport models come full circle. Cell 67:453–456, 1991. C, From Wang L, Brown A: Rapid intermittent movement of axonal neurofilaments observed by fluorescence photobleaching. Mol Biol Cell 12:3257–3267, 2001. Reprinted from Molecular Biology of the Cell [12:3257–3267, 2001] with the permission of The American Society for Cell Biology.)

Defining the mechanism of slow transport was challenging because various experimental approaches yielded apparently conflicting results. Radioactive labeling established the existence of slow movements and showed that the moving proteins become spread out and diluted as they move away from the cell body (Fig. 37-5A). Photobleaching of fluorescent tubulin and actin in axons of cultured neurons demonstrated that the bulk of these cytoskeletal polymers are stationary (Fig. 37-5B), whereas fluorescent tubulin, photoactivated inside an axon of a cultured Xenopus neuron, moves as a block at a rate that is characteristic of slow transport.

This puzzle was resolved by imaging single fluorescent intermediate filaments in axons of live nerve cells. These filaments are stationary most of the time (up to 99%), but occasionally, they move rapidly (0.2 to 2 mm s⁻¹) for up to 20 mm. Most but not all of these movements are away from the cell body, accounting for the net anterograde movement. These rapid but intermittent movements depend on microtubules and are presumably driven by motor proteins. The movements of whole microtubules are similar to those of intermediate filaments. Thus, fast but intermittent transport appears to be slow in assays for bulk transport.

Other Microtubule-Dependent Movements

Other cells use the same molecular mechanisms as neurons to move organelles in the cytoplasm. Secretory vesicles use plus-end motors to move from the Golgi apparatus to the plasma membrane. Endosomes use dynein to move from the plasma membrane toward the cell center. Herpes virus and rabies virus also use dynein for long-distance transport on microtubules from the terminals of sensory nerves to the cell body, where viral DNA enters the nucleus for replication.

The distribution of the endoplasmic reticulum de-pends on intact microtubules. Strands of the endoplasmic reticulum align with microtubules in cultured cells (Fig. 37-6). This codistribution is achieved in two ways: (1) Motors transport strands of endoplasmic reticulum bidirectionally on microtubules, and (2) other strands of the endoplasmic reticulum attach to the plus end of microtubules and ride the microtubule tip as it grows and shrinks during dynamic instability. This is the best example of movement of an organelle driven by microtubule assembly. The concentration of the Golgi apparatus near the centrosome depends on microtubules, because dynein motors transport Golgi vesicles toward the minus ends of the microtubules. Mitochondria move bidirectionally on microtubules in animal cells but depend on actin filaments in yeast. These examples illustrate how not only the dynamics of the organelles but also the overall organization of a cell depend on the activity of microtubule motors. Thus, cellular architecture is determined actively, not passively.

Figure 37-6 two modes of microtubule-dependent movement of the endoplasmic reticulum in a newt epithelial cell. The cell was microinjected with rhodamine-labeled tubulin, which incorporates into microtubules, and a lipophilic fluorescent dye (DiOC₆), which labels endoplasmic reticulum (ER). Time series are indicated in minutes and seconds. Scale bar for all panels is 5 mM. A, This column of fluorescence micrographs illustrates the dynamics of microtubules (red) and ER (green), over a period of 19 minutes. Note the strand of ER moving away from the leading edge (arrowheads). B, Time course of the movement of a strand of ER toward the end of a microtubule, followed by retraction. This type of movement is thought to be driven by a kinesin motor attached to the tip of the elongating membrane (arrowhead). C, Time course of the movement of a strand of ER attached to the tip of a growing microtubule (arrowhead), followed by retraction of the membrane along the microtubule.

(Courtesy of C. Waterman-Storer and E. D. Salmon, University of North Carolina, Chapel Hill. Reference: Waterman-Storer C, Salmon ED: Endoplasmic reticulum membrane tubules. Curr Biol 8:798–806, 1998.)

The cellular distribution of nucleoprotein complexes in the cell also depends on active movements. The most obvious example is the movement of chromosomes during mitosis, which depends on microtubule assembly and microtubule motors (see Fig. 44-7). Another example is the asymmetric localization in fly oocytes of certain mRNAs that help to establish the polarity of the embryos. Specific RNA sequences promote the assembly of protein-containing particles that move at steady rates on microtubules over long distances in the cell (Fig. 37-7). Kinesins and dynein are believed to power these movements, but most of the details remain to be determined. Dynein also anchors localized RNAs in oocytes.

Figure 37-7 Transport of mRNA for myelin basic protein in a cultured oligodendrocyte, a glial cell isolated from brain. mRNA synthesized in the cell body (or, in this case, labeled with a fluorescent dye and microinjected into the cell body) is packaged with proteins in a ribonucleoprotein particle, transported from the cell center along microtubules at a steady rate of 0.2 mm s⁻¹, and released at the periphery, where it moves randomly at 1 mm/s.

(Redrawn from Ainger K, Avossa D, Morgan F, et al: Transport and localization of exogenous myelin basic protein mRNA microinjected into oligodendrocytes. J Cell Biol 123:431–441, 1993, by copyright permission of The Rockefeller University Press.)

Intracellular Movements Driven by Microtubule Polymerization

Microtubule polymerization and depolymerization have long been known to play a central role in the assembly of the mitotic apparatus and the movement of chromosomes (see Fig. 44-7), as well as the establishment of cellular asymmetry (see Fig. 34-2). Polymerizing microtubules can exert substantial forces, but the force will buckle microtubules longer than about 10 mm. Consequently, microtubule pushing mechanisms work best over short distances, such as for positioning the nucleus in fission yeast cells and the mitotic spindle in budding yeast cells.

Remarkably, the depolymerizing end of a micro-tubule can also pull on attached cargo. An in vitro proof-of-principle experiment (Fig. 37-8) showed that chromosomes can ride along on the end of a depolymerizing microtubule, even in the absence of ATP or GTP. Presumably, multiple weak bonds between the chromosome and the side of the microtubule near its end rearrange rapidly enough to maintain attachment, even as tubulin subunits dissociate from the end. In budding yeast, the link between the chromosome and the end of the microtubule is a ring-shaped complex of ten proteins. This ring may slide along the microtubule as protofilaments peel away from the end as it shortens.

Figure 37-8 transport of an isolated chromosome on a shortening microtubule in vitro. A microtubule was grown from brain tubulin nucleated by a basal body in the extracted carcass of a ciliate, Tetrahymena. A chromosome (arrow) was added as a test cargo and captured by the end of the microtubule. When the concentration of tubulin was reduced, the microtubule shortened, carrying along the chromosome attached to its tip. This transport occurs in the absence of ATP or GTP.

(Courtesy of J. R. McIntosh, University of Colorado, Boulder.)

Endoplasmic reticulum provides the best example of an intracellular organelle that harnesses microtubule growth for movement as an alternative to motor-driven movements along microtubules (Fig. 37-6). A “tip attachment complex,” yet to be characterized, maintains a connection between the endoplasmic reticulum and the end of a microtubule as its length varies secondary to cycles of polymerization and depolymerization.

Bulk Movement of Cytoplasm Driven by Actin and Myosin

Bulk streaming of cytoplasm is most spectacular in plant cells (Fig. 37-1A). Although confined within rigid walls, plant cell cytoplasm streams vigorously at very high velocities (up to 60 mm s⁻¹). At this rate, cytoplasm moves 5 m/day. Such cytoplasmic streaming is best understood in the giant cells of the green alga Nitella. Streaming occurs continuously in a thin layer of cytoplasm between the large central vacuole and chloroplasts immobilized in the cortex. On each side of the cell, a zone of stationary cytoplasm separates streams moving in opposite directions. The physiological function of this streaming is not clearly understood.

Bulk streaming in Nitella is brought about by movement of endoplasmic reticulum along tracks consisting of bundles of polarized actin filaments associated with chloroplasts (Fig. 37-9C). All of the actin filaments in these bundles have the same polarity, and cytoplasm streams toward their barbed ends. In Nitella extracts, membrane vesicles move along actin filament bundles at the same high velocities that are characteristic of the cytoplasmic streaming. An extraordinary type XI myosin pulls endoplasmic reticulum along cortical actin tracks, dragging along other cytoplasmic components, including organelles and soluble molecules. This myosin moves nearly 10 times faster than the fastest muscle contraction, apparently by taking large steps and by the cooperation of several motors working rapidly on the same membrane.

Figure 37-9 cytoplasmic streaming in the green alga nitella. A, A pair of differential interference contrast (DIC) light micrographs showing the movement of organelles in cytoplasm. Note the strand of endoplasmic reticulum (ER [arrow]). B, Time series of DIC light micrographs showing movement of a vesicle isolated from Nitella along a bundle of actin filaments isolated from Nitella. C, Scanning electron micrographs of the cortex isolated from Nitella showing the bundles of actin filaments associated with chloroplasts. D, Transmission electron micrographs of a freeze-fracture preparation (upper) and thin section (lower) showing ER associated with actin filament bundles. E, Freeze-fracture preparation of a vesicle associated with an actin filament bundle. F, Movement of ER along actin filament bundles dragging along bulk cytoplasm.

(Courtesy of B. Kachar, National Institutes of Health. Reference: Kachar B, Reese T: The mechanism of cytoplasmic streaming in characean algal cells. J Cell Biol 106:1545–1552, 1988.)

A completely different actomyosin mechanism produces equally spectacular cytoplasmic streaming in the acellular slime mold Physarum. In these giant, multinucleated cells, cytoplasm flows back and forth rhythmically at high velocities through tubular channels (Fig. 37-10). Cycles of contraction and relaxation of cortical actin filament networks push the relatively fluid endoplasm back and forth in a manner akin to squeezing a toothpaste tube. Myosin-II is thought to generate the cortical contraction, as it is present in high concentration in this cell and can contract actin filament gels in vitro. (This, incidentally, was the first nonmuscle myosin to be purified in the late 1960s.) The cortical contractions that are so prominent in Physarum are also used by giant amoebas for cell locomotion (see Fig. 38-1), cytokinesis (see Fig. 44-23), and movements of some embryonic tissues (see Fig. 38-5).

Figure 37-10 cytoplasmic streaming in the acellular slime mold physarum polycephalum. A, Photograph of Physarum, a giant multinucleated single cell growing in a baking dish. B, Blur photomicrograph made with polarization optics by taking a time exposure showing the bulk streaming of the endoplasm in a cytoplasmic strand (long arrow). Mucus is designated by “M.” C, Time course of pressure changes produced by shuttle streaming of cytoplasm through a strand.

(B, From Nakajima H: The mechanochemical system behind streaming in Physarum. In Allen RD, Kamiya N (eds): Primitive Motile Systems in Cell Biology. New York, Academic Press, 1964, pp 111–123. C, Reference: Kamiya N: The mechanism of cytoplasmic movement in a myxomycete plasmodium. Symp Soc Exp Biol 22:199–214, 1968.)

Actin-Based Movements of Organelles in Other Cells

Like Nitella, budding yeasts transport vesicles along bundles of actin filaments from the mother to the bud (Fig. 37-11), although the movements of these solitary vesicles do not produce cytoplasmic streaming. Myosin-V is the motor, so vesicles fail to move from mother to bud in null mutants of myosin-V genes. A myosin-V also transports certain mRNAs along actin filament cables from the mother to the bud, where they determine cell fate.

Figure 37-11 Fluorescence micrographs (A–D) showing actin filament bundles and patches at various stages in the cell cycle of the budding yeast Saccharomyces cerevisiae. Myosin-V uses these actin filament bundles to deliver vesicles (including the vacuole), certain mRNAs, and at least one enzyme (chitin synthase) from the mother to the bud.

(Courtesy of J. A. Cooper, Washington University, St. Louis, Missouri.)

Animal cells generally use extended microtubules for long-distance movements and shorter actin filaments for local transport. For example, fish skin cells can change color by using dynein to aggregate and kinesin to disperse pigment granules called melanophores along radial microtubule tracks. Myosin-V contributes by moving dispersed melanophores laterally between microtubules. Similarly, mutations causing light coat color in mice revealed that myosin-V is required for some aspects of the transport of pigment granules called melanosomes within and between cells in the skin. A protein called melanophilin links the tail of myosin-V to a small GTPase on melanosomes.

Cytoplasmic Movements Driven by Actin Polymerization

Some intracellular pathogenic bacteria, including Listeria and Shigella, use actin polymerization to move through the cytoplasm of their animal cell hosts at about 0.5 mm s⁻¹ (Fig. 37-12A). These bacteria hijack the machinery that is used to move the leading edge of motile cells to polymerize a comet tail of actin filaments that pushes the bacterium forward. One end of the bacterium has a concentration of proteins that directly (Listeria) or indirectly (Shigella) activate Arp2/3 complex to polymerize a network of branched actin filaments (see Fig. 33-13). Growth of this network pushes the bacterial cell forward. The comet tail of cross-linked actin filaments is stationary and depolymerizes distally at the same rate at which it grows next to the bacterium, so it remains a constant length. Under some conditions, cellular endosomes can induce actin filament comets and move similar to Listeria rather than using microtubules. It is not yet known how widely this phenomenon is used for intracellular motility.

Figure 37-12 Fluorescence micrographs of actin filament comet tails in animal epithelial cells infected with the bacterium Listeria monocytogenes (A) or Vaccinia virus (B). Both pathogens are stained green with fluorescent antibodies. They use host cell proteins to assemble a cross-linked network of actin filaments shaped like a comet tail. Actin filaments are stained red with rhodamine-phalloidin. A, The comet tail pushes Listeria in a PtK cell through the cytoplasm and into projections of the plasma membrane at the edge of the cell. B, When the replicated Vaccinia viruses in this HeLa cell reach the cell surface 8 hours after infection, they activate Arp2/3 complex to assemble a cytoplasmic comet tail of actin filaments that are thought to enhance the spread of the virus from cell to cell. Actin-based motility of Vaccinia virus depends on tyrosine phosphorylation of a viral transmembrane protein A36R that remains inserted in the plasma membrane.

(A, Courtesy of K. Skoble, D. Portnoy, and M. Welch, University of California, Berkeley. B, Courtesy of T. P. Newsome and M. Way, Cancer Research UK, London, England. Reference: Frischknecht F, Moreau V, Rottger S, et al: Actin-based motility of Vaccinia virus mimics receptor tyrosine kinase signaling. Nature 401:926–929, 1999.)

Vaccinia viruses attached to the outer surface of animal cells also use transmembrane proteins to usurp the cytoplasmic actin assembly system to drive their movements at one stage in its life cycle (Fig. 37-12B). Placement of a plastic bead coated with adhesion proteins on the plasma membrane of some animal cells can induce similar propulsive actin comet tails in the cytoplasm.

Fungal and animal cells use Arp2/3 complex to assemble small comets of actin filaments associated with endocytic vesicles. Mutations show that endocytosis in yeast depends on actin assembly, but questions remain about the contribution of actin to the various steps in endocytosis (vesicle invagination, fission of the vesicle from the plasma membrane and movement of the vesicle from the plasma membrane), which is still being investigated.

ACKNOWLEDGMENT

Thanks go to Larry Goldstein for his suggestions on revisions to this chapter.