| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
*
* Department of Physics, Massachusetts Institute of Technology, Cambridge, Massachusetts; and Department of Biological Sciences, Columbia University, New York, New York
Correspondence: Address reprint requests to Michael P. Sheetz, Dept. of Biological Sciences, PO Box 2408, Columbia University, Sherman Fairchild Center, Rm. 713, 1212 Amsterdam Ave., New York, NY 10027. Tel.: 212-854-4857, Lab 4-8133; Fax: 212-854-6399; E-mail: ms2001{at}columbia.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
). Because biological membranes are fluid, the energy needed to move large amounts of membrane from one region to another is quite small (Dai and Sheetz, 1995b). Small tension gradients in the plane of the membrane are sufficient to power very rapid movements. We provide evidence here that the physical properties of Golgi and ER membrane tubules are particularly different and could provide a simple means for driving some of the movements from one compartment to another.
In the processes of protein synthesis and maturation, the initial events in the ER are followed by considerable bidirectional traffic between the ER and Golgi, the Golgi and plasma membrane, or the Golgi and lysosomes. Once a compartment is formed, communication with other compartments is limited and Golgi tubules going to the plasma membrane rarely fuse with ER or other internal membranes. The limited fusion of compartments enables them to maintain their specialized functions and to only modify the specific proteins that traffic through them. When the barriers to fusion do break down, the compartments can mix rapidly. For example, the drug Brefeldin A (BFA) causes Golgi membrane to fuse and co-mingle with ER, and has been used extensively to study mechanisms regulating membrane traffic. When a cell is treated with BFA, the Golgi membrane starts tubulating along microtubules and is rapidly drawn into the ER (Lippincott-Schwartz et al., 1989, 1990). Although intact Golgi tubules can extend through the cytoplasm for 5–10 min normally, the fusion of Golgi tubules with ER causes the Golgi contents to mix into the ER within 15–30 s (Sciaky et al., 1997). Sciaky and co-workers interpret their results as an indication of tension-driven flow between the two membrane compartments. The ER is thought to provide a lower energy environment for membrane protein and lipid than the Golgi system, and flow rather than diffusion could cause mixing of components. The first step in verifying this hypothesis is to measure the tensions in the ER and Golgi tubules directly.
A well-characterized measure of tension within membrane bilayers is the force on membrane tethers pulled from those membranes (reviewed in Sheetz, 2001; Morris and Homann, 2001). Tether force can be rapidly measured with laser tweezers using beads attached to membranes to form tethers. In the case of plasma membranes, tether force is inversely related to the rates of endocytosis, membrane resealing, and lamellipodial extension. Internal membranes are also under tension and differences in tension could drive movements in a manner consistent with observed events in vivo. By bringing membrane tubules into contact with each other using laser tweezers, it is possible to measure the difference between homotypic and heterotypic fusion.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
) with 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide (EDC). Briefly, 0.5 ml of a 2.5% suspension of carboxylate beads were washed in carbonate and phosphate buffers and incubated with a 2% EDC solution for 3–4 h at room temperature. The carbodiimide was washed with a phosphate buffer and the beads were resuspended in a borate buffer. After activation, the beads were coated with 1), anti-kinectin antibody, which has a preference for ER membranes; or 2), Wheat Germ Agglutinin (WGA), which binds preferentially to the Golgi membrane; or 3), Bovine Serum Albumin (BSA) as a control. 200–300 micrograms of the protein was added to the bead suspension and incubated overnight at 4°C with gentle end-to-end mixing. Beads were centrifuged and mixed with first ethanolamine and then BSA to block nonspecific protein binding sites. The beads were suspended in a storage buffer (containing sodium phosphate and sodium chloride) and stored at 4°C.
Formation of networks from membrane extracts
Internal membranes and cytosol rich in motors were obtained from chicken embryo fibroblasts following standard protocol (McIlvain et al., 1993). To separate the Golgi from the ER, the supernatant from a 1000-g spin of the cell homogenate was diluted with PMEE' + (standard buffer, McIlvain et al., 1993) to 1 ml and centrifuged at low speed (10,000 g) to obtain a pellet of the heavier membranes (H-fraction). The resulting supernatant was spun at high speed (100,000 g) to get a pellet of lighter vesicles (L-fraction). H- and L-fractions were then resuspended in PMEE' + and used for network formation and fluorescence assays. Ten-microliter capacity flow chambers were assembled from two coverslips separated by parallel strips of 70-µm-thick double-stick tape. The chambers were perfused with taxol-stabilized, bovine brain microtubules (10 µl, at 0.1–0.5 mg/ml), and incubated in a humid chamber for 15–20 min. Unbound microtubules were removed with 40 µl of washing buffer (PMEE' +, 1 mM GTP, 20 µM taxol). Membrane fractions (5 µl) with motor supernatant (3 µl) and Mg-ATP (2 µl) were introduced into the flow chambers and network formation was assayed after 
60 min incubation at 37°C.
Microscopy and force measurement
Network formation was imaged using video-enhanced differential interference contrast (DIC) microscopy as described before in Dabora and Sheetz (1988). Fluorescent images were taken using a cooled charge-coupled device camera (Princeton Instruments, Princeton, NJ). The force of the optical trap was calculated by computing the viscous drag of a bead through the aqueous medium as described before in Dai and Sheetz (1995a). A linear force-displacement graph was obtained to calculate the calibration constant for the trap stiffness. Carboxylate beads of 0.5-µm diameter (Polysciences) were covalently coupled (using user-supplied protocol) with either anti-kinectin antibody (which has a preference for ER membranes) or WGA (which binds preferentially to the Golgi membrane). Protein-coated beads were flowed into the chamber along with motor-containing supernatant with GTP, taxol, and Mg-ATP. Force measurement is described in Fig. 3. The recorded sequences of tether pulling events were digitized and analyzed using the tracking software ISEE (Inovision, Durham, NC)—a nanometer-level particle-tracking program that calculates the centroid of beads with a maximum precision of a few nanometers.
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
; Lane and Allan, 1999) and Golgi (Allan and Vale, 1991; Fullerton et al., 1998) membranes on microtubules. The dynamics and morphology of these reconstituted networks is remarkably similar to that in living cells (Lee and Chen 1988; Terasaki et al., 1986). They, therefore, provide an extremely useful cell-free system to conduct mechanical measurements on intracellular membranes.
In the process of forming networks, the membrane tubules were extremely dynamic on microtubules. An amorphous aggregate of membrane, which adhered to a microtubule meshwork on the glass coverslip, was the precursor of the network. Microtubule-dependent motors (kinesin and dynein) attached to regions of the membrane and moved along stationary microtubules, providing the force to draw out tubular branches. In the absence of motors or microtubules, we did not observe any tubular extensions, suggesting that motor force itself created tubules from a membrane with no preferred curvature, as opposed to motor proteins simply guiding pre-existing membrane extensions. The growing tubule was frequently observed to retract to its point of origin (possibly due to detachment from the microtubule). Membrane branches formed when another active motor contacted and moved along an intersecting microtubule, pulling a new membrane tubule from a pre-existing branch. A growing membrane branch fused with another branch if they overlapped. After fusion, the branches relaxed to a configuration connected by trigonal vertices with 120° angles between branches to minimize the local energy. This resulted in a reticular network of long membrane tubules on a microtubule mesh. In some instances, pre-existing polygons shrank in size due to movement of one of the branches and even disappeared, causing a local rearrangement of the network. After a few hours, the dynamics of tubules ceased and the entire structure stabilized. The network was interconnected and could stretch unbroken over hundreds of microns similar to the structures found in vivo (Terasaki et al., 1986). The tubulovesicular structure was attached to the underlying bed of microtubules at discrete points. The cause of attachment is yet unknown and could be due to inactive motors or some other attachment proteins. A typical example of a tubulovesicular network is shown in Fig. 1.
|
|
30 measurements for each case). The distribution of forces was normal. For all of these measurements, the identity of the ER and the Golgi membranes was confirmed by immunostaining.
The energy required to pull a tether is given by Bozic et al. (2001) and Bukman et al. (1996) as
 | (1) |
). The pressure difference term can be neglected because the tether volume is not conserved. Both ER and Golgi membranes do not tubulate spontaneously and require the action of motors or cytosolic factors to form tubular networks. Further, purified ER membranes tend to aggregate into large micron-sized vesicles (Dreier and Rapoport, 2000). These observations suggest that we can neglect the effect of spontaneous curvature. Taking these simplifications into account, the free energy is minimized with respect to the membrane shape for a cylinder (Bozic et al., 2001). An inverse relationship between tether (tubule) diameter and membrane tension indicates that the Golgi tubules should be larger in diameter than ER tubules. The tether force F is related to the radius of the tether, Rt, the bending stiffness of the membrane, and the membrane tension by the following equations (Hochmuth et al., 1996; Waugh et al., 1992):
 | (2) |
 | (3) |
 | (4) |
From the contrast of the tubules in the DIC images it was possible to estimate the relative diameters (Fig. 4). Golgi tubules had an average diameter of 180 nm whereas the ER network branches were 115 nm in diameter or a factor of RG/RE = 1.57 ± 0.2 (Fig. 4). The inverse ratio of tether forces is FE/FG = 1.63 ± 0.3. We find that the two ratios (FE/FG and RG/RE) are the same within experimental error. This implies that both membrane types have approximately the same bending stiffness: BE 
BG 3.3 x 10–19 N/m, similar to that of growth cone membranes and phospholipid bilayers (Dai and Sheetz, 1995a, Evans and Rawicz, 1990; Song and Waugh, 1993). Calculating the tensions explicitly using Eq. 3, we find that the tension in the ER membrane networks is TE 0.013 dyn/cm and the tension in the Golgi membrane is TG 0.005 dyn/cm. To put these tensions in perspective, membrane lysis requires a tension of 5–10 dyn/cm. As a further indication that the Golgi networks were at a lower tension, the Golgi tubules were more "floppy" (had larger thermal fluctuations) than the ER tubules. For this analysis we have assumed that the spontaneous curvature of the membranes is negligible. However, it is possible that both ER and Golgi have nonzero spontaneous curvatures. Therefore, the numerical values of tension and bending stiffness could be different from that reported here. Further work measuring the spontaneous curvatures of the two membrane types is required to determine more accurately the relative contributions of spontaneous curvature and tension to the tether force.
|
Because the membrane fractions were not pure we worried that fusion between ER and Golgi tubules could produce hybrid networks in vitro although the two compartments do not normally fuse in vivo. To study fusion, tethers were pulled from tubules, placed over a nearby network branch, and allowed to fuse for several seconds. For homologous membrane types (i.e., ER-ER or Golgi-Golgi), fusion was observed 70% of the time. For heterologous membrane types, ER to Golgi or vice versa, no fusion event was observed (25 trials). The behavior of membrane tubules in vitro is consistent with the properties of ER and Golgi in vivo. The much lower rate of heterologous versus homologous fusion of the networks in vitro helps to explain how separate compartments could be maintained in vitro as well as within the cell. Homologous fusion would explain how a reticular network of ER could form. If agents such as BFA would cause heterologous fusion, then the rapid movement of the Golgi into the ER would be explained by the greater tension in the ER.
Is the observed tension difference sufficient to cause membrane flow from Golgi to ER as observed after addition of BFA (Lippincott-Schwartz et al., 1989, 1990)? From the measured redistribution of a fluorescent Golgi protein into the ER, Sciaky et al. (1997) have concluded that movement of membrane protein between Golgi and ER is due to convective flow rather than diffusive movement, with velocities on the order of 10 µm/s. Chizmadzhev et al. (1999) have calculated the velocity v of lipid transfer between fusing membranes at different tensions to be v = C(
/
), where is the tension difference, is the surface viscosity of the membrane, and C is a factor depending on the pore geometry. Making reasonable assumptions about the pore geometry and viscosity of these membranes, = 10–5–10–6 g/s (Evans and Hochmuth, 1978; Saffman, 1976), we find that the measured tension difference (0.01 dyn/cm) is sufficient to produce the observed lipid flow velocities. Experiments on plasma membranes show that even the presence of a cytoskeleton does not increase the surface viscosity sufficiently to block this effect (Hochmuth et al., 1996). One possible origin of the tension is the action of microtubule motors that extend the membrane. Since motor inhibitors block network spreading but have no effect on the tether force of existing networks, it appears that there is a large reservoir of membrane at a constant tension. Further, throughout a single type of network we observe the same tubule radius, suggesting that the equilibrium radius is not set by the motor force, which could vary across the network, but by a constant chemical potential. The difference in tension could arise from a difference in chemical potential of the two membrane compartments. The cell may maintain the surface tension of its different membrane compartments at fixed levels by keeping lipid reservoirs at a fixed chemical potential. The normal block to heterologous fusion would maintain the separation between the different membrane types. Thus, the in vitro behavior of the tubulovesicular networks formed from ER and Golgi membrane provides important insights into aspects of function in vivo. Even though there could be differences between in vitro and in vivo situations, our experiments are a first step in identifying the physical properties of intracellular organelles. Experiments with BFA in vivo (Sciaky et al., 1997) show rapid movement of Golgi into the ER, ruling out diffusional mixing of the two compartments, which would be the case if the membranes were at the same tension. This is important evidence in favor of our hypothesis that tension differences measured in vitro could in fact reflect the in vivo situation. Differences in membrane tension between ER and Golgi provide a simple explanation for the rapid transfer of Golgi to ER.
Submitted on June 18, 2003; accepted for publication January 5, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Bozic, B., V. Heinrich, S. Svetina, and B. Zeks. 2001. Shapes of nearly cylindrical, asymmetric bilayer membranes. Eur. Phys. J. E. 6:91–98.
Bukman, D. J., J. H. Yao, and M. Wortis. 1996. Stability of cylindrical vesicles under axial tension. Phys. Rev. E. 54:5463–5468.
Buton, X., G. Morrot, P. Fellmann, and M. Seigneuret. 1996. Ultrafast glycerophospholipid-selective transbilayer motion mediated by a protein in the endoplasmic reticulum membrane. J. Biol. Chem. 271:6651–6657.
Chizmadzhev, Y. A., D. A. Kumenko, P. I. Kuzmin, L. V. Chernomordik, J. Zimmerberg, and F. S. Cohen. 1999. Lipid flow through fusion pores connecting membranes of different tensions. Biophys. J. 76:2951–2965.
Dabora, S. L., and M. P. Sheetz. 1988. The microtubule-dependent formation of a tubulovesicular network with characteristics of the ER from cultured cell extracts. Cell. 54:27–35.[Medline]
Dai, J., and M. P. Sheetz. 1995a. Axon membrane flows from the growth cone to the cell body. Cell. 83:693–701.[Medline]
Dai, J., and M. P. Sheetz. 1995b. Mechanical properties of neuronal growth cone membranes studied by tether formation with laser optical tweezers. Biophys. J. 68:988–996.
Dreier, L., and T. A. Rapoport. 2000. In vitro formation of the endoplasmic reticulum occurs independently of microtubules by a controlled fusion reaction. J. Cell Biol. 148:883–898.
Evans, E. A., and R. M. Hochmuth. 1978. Mechanochemical properties of membranes. Curr. Top. Membr. Transp. 10:1–62.
Evans, E. A., and W. Rawicz. 1990. Entropy-driven tension and bending elasticity in condensed fluid membranes. Phys. Rev. Lett. 64:2094–2097.[Medline]
Fullerton, A. T., M. Y. Bau, P. A. Conrad, and G. S. Bloom. 1998. In vitro reconstitution of microtubule plus end-directed, GTP
S-sensitive motility of Golgi membranes. Mol. Biol. Cell. 9:2699–2714.
Hochmuth, R. M., J. Y. Shao, J. Dai, and M. P. Sheetz. 1996. Deformation and flow of membranes into tethers extracted from neuronal growth cones. Biophys. J. 70:358–369.
Kuo, S. C., and M. P. Sheetz. 1993. Force of single kinesin molecules measured with optical tweezers. Science. 260:232–234.
Lane, J. D., and V. J. Allan. 1999. Microtubule-based endoplasmic reticulum motility in Xenopus laevis: activation of membrane-associated kinesin during development. Mol. Biol. Cell. 10:1909–1922.
Lee, C., and L. B. Chen. 1988. Dynamic behavior of endoplasmic reticulum in living cells. Cell. 54:37–46.[Medline]
Lippincott-Schwartz, J., L. C. Yuan, J. S. Bonifacino, and R. D. Klausner. 1989. Rapid redistribution of Golgi proteins into the ER in cells treated with Brefeldin A: evidence for membrane cycling from Golgi to ER. Cell. 56:801–813.[Medline]
Lippincott-Schwartz, J., J. G. Donaldson, A. Schweizer, E. G. Berger, H. Hauri, L. C. Yuan, and R. D. Klausner. 1990. Microtubule-dependent retrograde transport of proteins into the ER in the presence of Brefeldin A suggests an ER recycling pathway. Cell. 60:821–836.[Medline]
McIlvain, J. M., C. Lamb, S. Dabora, and M. P. Sheetz. 1993. Microtubule motor-dependent formation of tubulovesicular networks from endoplasmic reticulum and Golgi membranes. Methods Cell Biol. 39:227–236.[Medline]
Morris, C. E., and N. Homann. 2001. Cell surface area regulation and membrane tension. J. Membr. Biol. 179:79–102.[Medline]
Saffman, P. G. 1976. Brownian motion in thin sheets of viscous fluid. J. Fluid. Mech. 73:593–602.
Schnapp, B. J., J. Gelles, and M. P. Sheetz. 1988. Nanometer-scale measurements using video-light microscope. Cell Motil. Cytoskeleton. 10:47–53.[Medline]
Sciaky, N., J. J. Presley, C. Smith, K. J. M. Zaal, N. Cole, J. E. Moreira, M. Terasaki, E. Siggia, and J. Lippincott-Schwartz. 1997. Golgi tubule traffic and the effects of Brefelden A visualized in living cells. J. Cell Biol. 139:1137–1155.
Sheetz, M. P. 2001. Cell control by membrane-cytoskeleton adhesion. Nat. Rev. Mol. Cell Biol. 2:392–396.[Medline]
Song, J., and R. E. Waugh. 1993. Bending rigidity of SOPC membrane containing cholesterol. Biophys. J. 64:1967–1970.
Terasaki, M., L. B. Chen, and K. Fujiwara. 1986. Microtubules and the endoplasmic reticulum are highly interdependent structures. J. Cell Biol. 103:1557–1568.
Waugh, R. E., J. Song, S. Svetina, and B. Zeks. 1992. Local and nonlocal curvature elasticity in bilayer membrane by tether formation from lecithin vesicles. Biophys. J. 61:974–982.
This article has been cited by other articles:
 |
|
 |
|
  E. Glassinger and R. M. Raphael Influence of Thermally Driven Surface Undulations on Tethers Formed from Bilayer Membranes Biophys. J., July 15, 2006; 91(2): 619 - 625. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Sun, J. S. Graham, B. Hegedus, F. Marga, Y. Zhang, G. Forgacs, and M. Grandbois Multiple Membrane Tethers Probed by Atomic Force Microscopy Biophys. J., December 1, 2005; 89(6): 4320 - 4329. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Cuvelier, I. Derenyi, P. Bassereau, and P. Nassoy Coalescence of Membrane Tethers: Experiments, Theory, and Applications Biophys. J., April 1, 2005; 88(4): 2714 - 2726. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Vedrenne, D. R. Klopfenstein, and H.-P. Hauri Phosphorylation Controls CLIMP-63-mediated Anchoring of the Endoplasmic Reticulum to Microtubules Mol. Biol. Cell, April 1, 2005; 16(4): 1928 - 1937. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Leduc, O. Campas, K. B. Zeldovich, A. Roux, P. Jolimaitre, L. Bourel-Bonnet, B. Goud, J.-F. Joanny, P. Bassereau, and J. Prost From the Cover: Cooperative extraction of membrane nanotubes by molecular motors PNAS, December 7, 2004; 101(49): 17096 - 17101. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
(c and d) Normalized intensity profile across a network branch from the Golgi (c) and ER (d). The lower contrast of the ER branch is indicated by the smaller area under the intensity profile curve.