This protocol describes a process for fabricating lipid nanotube networks using gliding kinesin motility in conjunction with giant unilamellar lipid vesicles.
Lipid nanotube (LNT) networks represent an in vitro model system for studying molecular transport and lipid biophysics with relevance to the ubiquitous lipid tubules found in eukaryotic cells. However, in vivo LNTs are highly non-equilibrium structures that require chemical energy and molecular motors to be assembled, maintained, and reorganized. Furthermore, the composition of in vivo LNTs is complex, comprising of multiple different lipid species. Typical methods to extrude LNTs are both time- and labor-intensive, and they require optical tweezers, microbeads, and micropipettes to forcibly pull nanotubes from giant lipid vesicles. Presented here is a protocol for the gliding motility assay (GMA), in which large scale LNT networks are rapidly generated from giant unilamellar vesicles (GUVs) using kinesin-powered microtubule motility. Using this method, LNT networks are formed from a wide array of lipid formulations that mimic the complexity of biological LNTs, making them increasingly useful for in vitro studies of lipid biophysics and membrane-associated transport. Additionally, this method is capable of reliably producing LNT networks in a short time (<30 min) using commonly used laboratory equipment. LNT network characteristics such as length, width, and lipid partitioning are also tunable by altering the lipid composition of the GUVs used for fabricating the networks.
The fabrication of lipid nanotube (LNT) networks is of increasing interest for in vitro examination of nonequilibrium lipid structures1,2,3. Cells use lipid tubules for the diffusive transport of proteins4 and nucleic acids5 as well as cell-to-cell communication6,7. The endoplasmic reticulum and Golgi apparatus are particularly interesting, as these membrane-bound organelles are the primary locations for lipid and protein synthesis as well as transport of these integral biomolecules within the cytoplasm of a cell8,9. The membranes of these organelles are comprised of multiple lipid species including sphingolipids, cholesterol, and phospholipids10 that ultimately help define their functionality. Thus, to more closely replicate and study these organelles, in vitro LNTs must be fabricated from vesicles with increasingly complex lipid formulations11.
Giant unilamellar vesicles (GUVs) are used pervasively for studying lipid membrane behavior because they can be reliably synthesized with complex formulations that include cholesterol, phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS), and phosphatidylinositol (PI)12,13. Described here is a method to fabricate LNTs from GUVs with varying lipid formulations using the gliding motility assay (GMA), in which LNTs are extruded based on the work performed by kinesin motors and microtubule filaments acting on GUVs. In this system, kinesin motor proteins adsorbed to a surface propel biotinylated microtubules, converting chemical energy from the hydrolysis of ATP into useful work (specifically, the extrusion of LNTs from biotinylated vesicles)11. The resulting LNT network provides a model platform to study effects of the differences in lipid phases on changes in LNT morphology.
Briefly, kinesin motor proteins are introduced into a flow chamber in a solution containing casein, which enables the adsorption of the motors onto the glass surface of the chamber. Next, biotinylated microtubules in a solution containing ATP flow through the chamber and are allowed to bind to the kinesin motors and begin motility. A streptavidin solution is then introduced into the chamber and allowed to bind non-covalently to the microtubules. Finally, GUVs containing a biotinylated lipid are introduced into the chamber and bind to the streptavidin-coated microtubules, then extrude LNTs to form large-scale networks over the course of 15–30 min. This method produces large, branched LNT networks using standard laboratory equipment and reagents at a low cost11.
1. Preparation of stock microtubule solutions
CAUTION: Safety goggles, gloves and a lab coat should always be worn throughout the protocol.
2. Preparation of giant unilamellar vesicles (GUVs)
3. Preparation of motility assay stocks and reagents
4. Gliding motility assay (GMA)
5. LNT network characterization
LNT networks (Figure 4) were fabricated using the described protocol, which uses the work performed by kinesin transport of microtubules to extrude LNTs from GUVs. Briefly, GUVs were prepared using agarose gel rehydration using sucrose solution, and microtubules were polymerized in GPEM solution and stabilized in BRB80T. Next, kinesin motors were introduced into a flow cell forming an active layer of motors on the surface of the coverslip. Microtubules were then introduced and a streptavidin solution was added, which facilitated binding between the biotinylated lipids and GUVs (which were subsequently introduced).
With all components present in the flow cell, LNTs were allowed to form for 30 min, at which point AMP-PNP was introduced to stop motility. Then, the LNTs were imaged under an epifluorescence microscope using a 100x oil objective. LNTs can be quite large; thus, a lower powered objective may also be used to capture larger LNTs. The LNT networks are characterized by thin, web-like protrusions extending from and connecting GUVs. The number and branching of the LNTs is dependent on several factors, including the density of microtubules on the surface, streptavidin concentration, and number of GUVs present.
This method can generate GUVs and LNTs composed of solid, liquid-disordered, and liquid-ordered phases, as well as phase-separated vesicles that exhibit the coexistence of these phases over a wide range of different compositions (Figure 4). For instance, synthesizing vesicles with 45% saturated lipid and 55% unsaturated lipid results in vesicles that separate into coexisting liquid disordered and solid phases. If cholesterol is included, however, liquid-liquid coexistence can then be observed. For example, a mixture comprised of 50% unsaturated lipids, 30% cholesterol, and 20% saturated lipids will create GUVs that separate into coexisting liquid-ordered (Lo) and liquid-disordered (LD) phases. The incorporation of cholesterol in this formulation fluidizes the saturated lipids, enabling the formation of a liquid phase.
Moreover, nodes (i.e., round spherical structures larger than the LNTs) were observed to form in the phase separated mixtures (Figure 4). Partitioning of lipid types within nanotubes and nodes may be characterized using the line profile tool was to find the maximum background-subtracted peak intensity of a node. The line profiles of the node and the LNT were determined and the background fluorescence of these profiles was subtracted to determine the maximum value. Partitioning is then calculated by dividing the maximum value of the fluorescence in the node by the maximum fluorescence value of the LNT. This approach enables the partitioning of lipids in both the LNT as well as nodes that form in the larger networks.
Figure 1: Composite image of LNTs fabricated from GUVs and microtubules. LNTs are extruded from GUVs by motile microtubules gliding on top of kinesin motors. Scale bar = 10 µm. Please click here to view a larger version of this figure.
Figure 2: Thresholding process to acquire thickness. (A) Select Threshold under the Image | Adjust tab. (B) Apply the threshold. (C) Draw a rectangle of known length over the desired tube. Black pixels have a value of 0 and red pixels have a value of 255. (D) Measure the integrated density of the area. To calculate tube width, divide the integrated density by 255, then divide this output by the length of the rectangle created in (B). Please click here to view a larger version of this figure.
Figure 3: Determining the lipid partitioning in nodes. (A) Open the image in ImageJ and use the line tool to draw a line over the desired node. (B) Measure the node intensity in the Oregon Green channel. (C) Move the line to the LNT and measure the intensity in the Oregon Green channel. (D,E,F) Repeat process for Texas Red channel. Scale bar = 20 µm. Please click here to view a larger version of this figure.
Figure 4: LNTs fabricated from different GUV lipid formulations. LNTs extruded from the liquid disordered (LD) phase are thin and long, while LNTs extruded from the liquid ordered (Lo) phase are short and thick. LNTs of both types are observed when Lo-LD phase-separated vesicles are used in the GMA. LNTs from liquid-solid phase GUVs resemble those extruded from LD GUVs. Liquid-ordered formulations can be created using a combination of saturated lipids and cholesterol, while liquid-disordered formulations are created using unsaturated lipids. Scale bars = 20 µm. Please click here to view a larger version of this figure.
LNT networks are a useful tool for in vitro studies to membrane properties and the transport of biomolecules such as transmembrane proteins. Moreover, using complex lipid formulations to fabricate LNT networks enables more biologically relevant studies. Other fabrication studies have used either 1) simple lipid formulations and multilamellar vesicles or 2) more cumbersome motility techniques to fabricate networks from GUVs comprised of complex lipid formulations. The method described here enables the efficient fabrication of large-scale LNT networks from complex lipid formulations and GUVs and using inexpensive reagents and equipment. As such, this methodology offers the ability to study a range of biological processes, including phase separation and membrane protein transport. The protocol uses an in vitro model in which the composition of the LNTs may be optimally tuned to approximate their biological analogs (e.g., Golgi apparatus).
The formation of LNT networks in the method described here has numerous advantages. For example, microtubules are easily and rapidly polymerized using lyophilized tubulin, and they remain stable at RT for at least 1–2 weeks when stabilized with paclitaxel16. As such, they are easily implemented to drive the extrusion of LNTs and formation of a large-scale networks1,2. Additionally, GUVs composed of multiple lipid components (i.e., unsaturated and saturated lipids and cholesterol) can be quickly formed based on a previous protocol with slight modifications. These modifications enable the synthesis of the GUVs capable of undergoing the physical chemical process of membrane phase separation. Once prepared, GUVs can be stored for weeks to months depending on storage conditions. However, it is recommended to use GUVS for up to 1 month before preparing a new batch. Lipid solutions can be stored at -20 °C or -80 °C for several months, as well as the agarose gel and coated coverslips, which can be stored at RT for months. The lipid films on the agarose-coated coverslips must be stored under a vacuum and rehydrated within 48 h.
One challenge of using this technique to form LNTs from GUVs is balancing the concentration of GUVs, streptavidin, and microtubules in the flow cell. For example, LNT formation will be limited if the ratio between microtubules to GUVs is not correct. If concentration of GUVs is too low, aggregates will not form, which is the first step in LNT formation. For the concentrations of microtubule and GUV described in this protocol, a 10x dilution of microtubule stock and 12x dilution of GUV stock generally resulted in good LNT networks. Dilutions of 5x and 6x, respectively, have also yielded good networks.
Another challenge is ensuring that the GUV solution is osmotically balanced with the microtubule solution. If the difference in osmolarity between the two solutions is too large, the GUVs will become unstable and potentially burst. If the solutions are not osmotically balanced (e.g., a 10% difference between the measured osmolarity between the two solutions), then a concentrated (e.g., 2 M) sucrose solution should be used to increase the osmolarity of the solution with the lower measured osmolarity. Another limitation of this system is the stability of the resulting LNT networks, which are stable on the order of hours and highly dependent on the flow chamber being continuously hydrated (or sealed the chamber with sealant). Once the solution has evaporated from the flow chamber, the LNTs will no longer be useful, despite the fact that the presence of a few residual LNTs may persist after evaporation.
The LNT networks created from these gliding microtubules and GUVs may be useful in understanding lipid bilayer dynamics as well as protein (e.g., transmembrane proteins) diffusion on membrane surfaces. The protocol described here can quickly create LNT networks comprised of various lipid formulations that more readily mimic biological LNT-like tunneling nanotubes as well as nanotubes found in membrane-bound organelles (i.e., endoplasmic reticulum and Golgi apparatus). The ability to form large-scale LNT networks is a key first step towards studying cell communication, studying nanofluidic biomolecule transport, and developing synthetic neuronal networks. This protocol opens the door to broader studies on the physiochemical properties of LNTs using a minimal, in vitro model system in which the composition of the LNTs can be easily modified to mimic natural cellular structures.
The authors have nothing to disclose.
This work was supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences and Engineering (BES-MSE). Kinesin synthesis and fluorescence microscopy were performed through a user project (ZIM) at the Center for Integrated Nanotechnologies, an Office of Science User Facility operated for the U.S. Department of Energy (DOE) Office of Science.
100x/1.4 Numerical Aperture Oil Immersion Objective | Olympus | 1-U2B836 | Olympus UPlanSApo 100x/1.40 Oil Objective Infinity Corrected, RMS Thread Working Distance 0.12mm |
3.0 ND Filter | Olympus | Neutral Density Filter | |
AMP-PNP | Sigma-Aldrich | A2647 | (β,γ-imidoadenosine 5′-triphosphate) |
ATP | Sigma-Aldrich | A7699 | Adenosine 5'-triphosphate disodium salt hydrate BioXtra |
Brightline Pinkel DA/FI/TR/Cy5/Cy7-5X-A000 filter set | Semrock | LED-DA/FI/TR/Cy5/Cy7-5X-A-000 | BrightLine Pinkel filter set, optimized for DAPI, FITC, TRITC, Cy5 & Cy7 and other like fluorophores, illuminated with LED-based light sources |
Casein | Sigma-Aldrich | 22090 | Casein hydrolysate for microbiology |
Catalase | Sigma-Aldrich | C9322 | Catalase from Bovine Liver |
Chloroform | Sigma-Aldrich | 288306 | Chloroform anhydrous contains 0.5-1.0% ethanol as stabilizer |
Cholesterol | Avanti | 700000P | cholesterol (ovine wool, >98%) (powder) |
D-Glucose | Sigma-Aldrich | G7021 | D-(+)-Glucose powder, BioReagent, suitable for cell culture, suitable for insect cell culture, suitable for plant cell culture, ≥99.5% |
DOPC | Avanti | 850375C | 1,2-Dioleoyl-sn-glycero-3-phosphocholine (in chloroform) |
DOPE-Biotin | Avanti | 870282C | 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(biotinyl) (sodium salt) |
DPPC | Avanti | 850355P | 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (powder) |
DPPE-Biotin | Avanti | 870285P | 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(biotinyl) (sodium salt) |
DTT | Sigma-Aldrich | 43816 | DL-Dithiothreitol solution 1 M |
EGTA | Sigma-Aldrich | E4378 | EGTA, Egtazic acid, Ethylene-bis(oxyethylenenitrilo)tetraacetic acid, Glycol ether diamine tetraacetic acid |
Glucose Oxidase | Sigma-Aldrich | G6125 | Glucose Oxidase from Aspergillus niger Type II, ≥10,000 units/g solid (without added oxygen) |
Glycerol | Fisher | G33 | Glycerol (Certified ACS), Fisher Chemical |
GTP | Sigma-Aldrich | G8877 | Guanosine 5′-triphosphate sodium salt hydrate |
IX-81 Olympus Microscope | Olympus | N/A | IX81 Inverted Microscope from Olympus |
KOH | Sigma-Aldrich | 1050121000 | Potassium Hydroxide |
Magnesium Chloride | Sigma-Aldrich | M1028 | 1.00 M magnesium chloride solution |
Orca Flash 4.0 Digital Camera | Hamamatsu | C13440-20CU | ORCA-Flash 4.0 V3 Digital CMOS camera |
Oregon Green-DHPE | Invitrogen | O12650 | Oregon Green 488 1,2-Dihexadecanoyl-sn-Glycero-3-Phosphoethanolamine |
Paclitaxel | ThermoFisher | P3456 | Paclitaxel (Taxol Equivalent) – for use in research only |
PIPES | Sigma-Aldrich | P6757 | 1,4-Piperazinediethanesulfonic acid, Piperazine-1,4-bis(2-ethanesulfonic acid), Piperazine-N,N′-bis(2-ethanesulfonic acid) |
Texas Red-DHPE | Invitrogen | T1395MP | Texas Red 1,2-Dihexadecanoyl-sn-Glycero-3-Phosphoethanolamine, Triethylammonium Salt |
Trolox | Sigma-Aldrich | 238813 | (±)-6-Hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid |
Tubulin, Biotin | Cytoskeleton | T333P | Tubulin protein (biotin) porcine brain |
Tubulin, Hy-Lite 488 | Cytoskeleton | TL488M | Tubulin protein (fluorescent HiLyte 488) porcine brain |
Tubulin, Unlabeled | Cytoskeleton | T240 | Tubulin protein porcine brain |