You are here: Vale Lab Home Page>Research>Other Projects>Organelle Transport>Reconsituting Organelle Transport

Microtubule-Based Organelle Transport


Reconstituting Organelle Transport Using Dictyostelium discoideum


In higher eukaryotes, long distance transport appears to be predominantly microtubule-based, while actin-based membrane transport plays a role in short-range transport. In the last five years, several kinesin superfamily members and cytoplasic dynein have been shown to transport membrane cargo. In many cases, particular kinesin motors appear to transport quite specific types of membranes. Elucidating the mechanism by which motors dock onto specific membranes is a major challenge for the future.

Reconstituted assays have traditionally provided powerful tools for identifying and dissecting functions of molecules involved in a cellular process. In vitro assays for organelle transport have worked well for the squid giant axon (which can only be obtained in small quantities), but have been very difficult in other systems. Many organelle transport assays reported in the literature are not very robust (very few organelle observed moving), making them difficult to use for biochemical experiments. Nira Pollock (a former graduate student) developed conditions for obtaining very robust bidirectional transport of membrane organelles along microtubules in Dictyostelium extracts. She initially used this assay to demonstrate a dominant-negative inhibition of minus-end-directed transport by a fragment of the dynein heavy chain.

(Click on image for a larger view)

Dictyostelium vesicles moving along microtubules nucleated from sea urchin axonemes in vitro. Since longer microtubules emerge from the plus end of the axonemes, the polarity of organelle transport can be identified.


Nira then took on the difficult task of using the in vitro organelle transport assay as a means of fractionating Dictyostelium cytosol to identify factors that cause organelles to move along microtubules. For example, when column chromatography was performed, Nira combined each column fraction with salt-washed Dictyostelium membranes to see if they would stimulate organelle transport as observed by video DIC microscopy. When this project began, we were uncertain whether we would emerge with a motor protein, an activator of a motor protein, or a combination of factors. However, after a microtubule affinity step, a hydroxyapatitite column, and a monoQ column, Nira identified two fractions, each containing predominantly a single polypeptide, that independently stimulated organelle transport at frequencies and velocities similar to those observed in the crude extracts.


(Click on image for a larger view)

Purification table for organelle transport factors from Dictyostelium.


Upon microsequencing, both polypeptides were identified as members of the kinesin superfamily. One is a close homologue of conventional kinesin (unpublished results) and the other is a close homologue of Unc104/KIF1A motors. The Unc104 gene was initially uncovered in a genetic screen for uncoordinated or immotile worms. It was later discovered that Unc104 encodes a motor protein of the kinesin superfamily and that the Unc104 mutation (effectively a null mutation) fails to transport synaptic vesicle precursors out of the cell body and into the axons. Similarly, mice that are deficient in the mammalian homologue (called KIF1A) also fail to transport synaptic vesicles, and form normal synapses and have high neuronal cell death leading to lethality. Additional kinesins with similar motor domain sequences to Unc104/KIF1A have been also uncovered (e.g. KIF1B and KIF1C, etc), but they have distinct tail domains and biological functions. Our findings show that the Unc104/KIF1A motor is ancient and evolved for vesicle transport functions in unicellular organisms.

The Unc104/KIF1A motor is the predominant plus-end-directed microtubule motor in Dictyostelium extracts and therefore was chosen for further study. Based upon our purification data, we believe that this motors can interact with and transport membranes without any accessory factors. Unc104 (C. elegans) and KIF1A (mouse) are both monomeric. Interestingly, in contrast to the higher eukaryotic homologues, Dictyostelium Unc104 (DdUnc104) has an extended coiled-coil at its C-terminus and is a dimer by hydrodynamic studies. This finding indicates that this family of kinesin motors is not exclusively "monomeric" and raises the possibility that the synaptic vesicle Unc104/KIF1A motors may dimerize under some circumstances. (See our recent abstract submitted to the ASCB meeting)

Nira also investigated the in vivo role of DdUnc104 by knocking out the gene by homologous recombination. The DdUnc104 null cells grew normally and showed no gross morphological defects. However, video microscopy revealed a 60% reduction in organelle transport in these cells. Mitochondria (detected with MitoTracker) moved normally, indicating that the transport defect was specific. In cell extracts, where the polarity of organelle transport could be unambiguously defined, plus-end-directed transport was dramatically decreased (90%); the remainder of the activity is likely due to the conventional kinesin motor, although we were unsuccessful in obtaining a gene knockout of this motor to prove this hypothesis. The in vitro organelle transport assay and DdUnc104 knockout cells provide important tools for investigating the mechanism of organelle transport. (Click here for more info.)

Despite the identification of numerous kinesin motor family members involved in membrane transport, it has been unclear how an individual motor binds and transports cargo vesicles specifically. The search for putative membrane receptors for kinesin and kinesin-related proteins only recently identified several proteins with already known functions that specifically interact with one type of motor. In contrast, little is known how membrane lipids may participate in facilitating motility. In our present research, we discovered that a novel lipid-based cargo-binding mechanism of the monomeric kinesin motor Unc104. C. elegans Unc104 (KIF1A in mammals) is responsible for the synaptic vesicle transport in neurons. Studying the motor/cargo interaction in a reconstituted membrane transport assay revealed that Dictyostelium Unc104, DdUnc104 transports membranes, at least in part, through a direct lipid interaction (Klopfenstein et al., 2002). A most likely candidate cargo binding domain is the pleckstrin homology (PH) domain in the C-terminal part of Unc104 family members. PH domains bind specifically phosphorylated lipids of the phosphatidylinositol species. The PH domain of DdUnc104 binds preferentially to phosphatidylinositol-(4,5)-bisphosphate (PIP2). Interestingly, protein-free liposomes are efficiently transported suggesting that this novel motor-lipid interaction is a bona-fide mechanism for cargo binding. Surprisingly, liposome transport by Unc104 motors is triggered in a highly cooperative manner above a certain PIP2 concentration. Clustering of PIP2 into lipid rafts increases transport frequencies at lower PIP2 concentrations suggesting that a critical local motor concentration is required for transport. These results can be explained by a dimerization model in which the usually monomeric Unc104 is more likely to dimerize and thereby activated; a notion consistent with processive, artificial Unc104 dimers. (See our work on the biophysics of Unc104.) Thus, we suggest that concentration of Unc104 within PIP2 -containing lipid domains triggers motor dimerization, which then allows for processive vesicle motion.



(click to see movie)

Figure1a (MOVIE): U446-PH. Mini-motors (U446-PH) with a PH domain transport vesicles and liposomes in vitro. Velocities are consistant with the velocities of Unc104 in a microtubule gliding assay (plus-end directed, 1.5 um/s). When the PH is attchached to the minus-end directed motor, ncd, cargo is transported in the opposite direction (to the microtubule minus-end) with the directionality and velocity consistent with ncd in a gliding assay (minus-end directed, 0.17 um/s).

Click on image above to see movie. If movie does not work, click here.


(click to see movie)

Figure 1b (MOVIE): PH-ncd.

(Click on image for a larger view)

Figure 2: Cooperativity of activation of Unc104-induced motility of PIP-2 containing liposomes but not motor binding to liposomes. Cooperativity is shifted by motor clustering into cholesterol/sphingomyelin induced lipid raf domains.


(Click on image for a larger view)

Figure 3: A membrane cargo containing PtdIns(4,5)P2 lipids, transmembrane proteins, and bound Unc104 motors is shown. In the absence of sufficient PtdIns(4,5)P2 density or clustering, Unc104 motors are dispersed. Since the Unc104 motor is not processive, individual motors transiently interact with the microtubule, but this is insufficient to move the vesicle continuously along the microtubule. PtdIns(4,5)P2 organization into membrane rafts could activate transport:Unc104 is a monomer in solution, but the concentration of several motors in a cluster may cause dimerization via predicted coiled-coil regions adjacent to the motor domain. A single, dimerized motor then could move along the microtubule by a hand-over-hand mechanism and processively transport the vesicle.


updated 4/9/07

Vale Lab Research

back to Home Page