Copyright 2008 American Chemical Society
JULY 15, 2008 VOLUME 24, NUMBER 14
Letters Directing the Polar Organization of Microtubules Erik D. Spoerke,*,† George D. Bachand,‡ Jun Liu,†,§ Darryl Sasaki,‡,| and Bruce C. Bunker† Department of Electronic and Nanostructured Materials, Sandia National Laboratories, 1515 Eubank SE, Albuquerque, New Mexico 87123, and Department of Biomolecular Interfaces and Systems, Sandia National Laboratories, 1515 Eubank SE, Albuquerque, New Mexico 87123 ReceiVed February 15, 2008. ReVised Manuscript ReceiVed April 15, 2008 Microtubules (MTs) are polar protein filaments that participate in critical biological functions ranging from motor protein direction to coordination of chromosome separation during cell division. The effective facilitation of these processes, however, requires careful regulation of the polar orientation and spatial organization of the assembled MTs. We describe here an artificial approach to polar MT assembly that enables us to create three-dimensional polar-oriented synthetic microtubule organizing centers (POSMOCs). Utilizing engineered MT polymerization in concert with functionalized micro- and nanoscale particles, we demonstrate the controllable polar assembly of MTs into asters and the variations in aster structure determined by the interactions between the MTs and the functionalized organizing particles. Inspired by the aster-like form of biological structures such as centrosomes, these POSMOCs represent a key step toward replicating biology’s complex materials assembly machinery.
Microtubules (MTs) are polymeric protein filaments that, along with intermediate and actin filaments, compose the cytoskeleton of eukaryotic cells. MTs act in a number of critical capacities within a cell, including participating in chromosome separation during cell division, directing the migration of neuronal axons, and serving as directional tracks for motor protein-based intracellular cargo transport.1,2 All of these functions, however, are critically dependent on (1) an inherent biochemical polarity of the MTs, and (2) the careful organization and orientation of * Corresponding author. Address: Sandia National Laboratories, Department of Electronic and Nanostructured Materials, P.O. Box 5800, MS 1411, Albuquerque, NM 87185-1411. E-mail:
[email protected]. † Department of Electronic and Nanostructured Materials. ‡ Department of Biomolecular Interfaces and Systems. § Current address: Pacific Northwest National Laboratory, 902 Battelle Boulevard, K2-50, Box 999, Richland, Washington 99352. | Current address: Sandia National Laboratories, 7011 East Avenue, Livermore, California 94550.
(1) Hirokawa, N.; Takemura, R. Nat. ReV. Neurosci. 2005, 6, 201–214. (2) Lodish, H.; Berk, A.; Zipursky, S.; Matsudaira, P.; Baltimore, D.; Darnell, J. In Molecular Cell Biology, 4th ed.; W. H. Freeman and Co.: New York, 1999; pp 795-836.
these polar filaments. We introduce here an artificial approach to MT organization that directs the polar assembly of MTs into a three-dimensional architecture. The intrinsic polarity of these hollow cylinders (25 nm in diameter) is a consequence of their polymerization from alternating R- and β-tubulin, creating a β-terminated “plus” end, and an R-terminated “minus” end. By controlling the MT orientation during organization within a cell, biological systems establish an intrinsic method for controlling the directionality of certain cellular functions. For example, networks of polar MTs control bidirectional motor protein-based nanocargo in which N-terminal kinesins move toward the “plus” end, while dyneins and C-terminal kinesins move toward the “minus” end of the MT. By controlling the polar organization of the MTs, the directional motility of the cargo on these motors can be controlled. This biological assembly scheme is an attractive model for synthetic materials synthesis and assembly. MTs themselves have a number of traits that would make them valuable for materials assembly: they are monodisperse in diameter, their length can be controlled, and they are capable of the dynamic assembly and
10.1021/la800500c CCC: $40.75 2008 American Chemical Society Published on Web 06/20/2008
7040 Langmuir, Vol. 24, No. 14, 2008
Letters
Figure 1. Schematic illustration of multistage MT polymerization and assembly into “plus-in” POSMOCs.
disassembly that would enable reconfigurable assemblies. Furthermore, the inherently selective relationship MTs have with motor proteins makes them attractive as components in a dynamic motor protein-based transport scheme for nanocargo manipulation. In fact, kinesin motors have already been used to transport nano- and microscale cargo over MT arrays.3,4 One of the key challenges to effectively exploiting MTs for materials assembly, however, is the crucial organization of the MT networks that direct the nanocargo transport. To date, significant effort has been devoted to developing one-dimensional arrays of MTs on solid surfaces as a means for directing the transport of nanoscale cargo in bead-based assays.3–7 These systems, however, lack the three-dimensional organization necessary for complex, multidimensional nanocargo transport such as that commonly found in natural systems. For example, in animal cells, MTs are nucleated from a centralized centrosome or microtubule organizing center (MTOC) to create three-dimensional MT constructs within the cell. Previous efforts to create these three-dimensional MTOCs in vitro have involved MT nucleation on isolated centrosomes8–10 or growing MTs from randomly oriented MT seeds covalently bound to micron-sized spheres.11,12 Others have employed modified kinesin motor protein “constructs” that arrange preformed MTs to form elegant, densely packed asters.13–15 In each of these cases, the application of the assembled structure is limited by its polarity. The methods using centrosomes or microspheres facilitated the formation of assemblies where, if there existed (3) Bohm, K.; Beeg, J.; zu Horste, G.; Stracke, R.; Unger, E. IEEE Trans. AdV. Packag. 2005, 28(4), 571–576. (4) Limberis, L.; Madga, J. L.; Stewart, R. J. Nano Lett. 2001, 1(5), 277–280. (5) Bohm, K. J.; Stracke, R.; Muhlig, P.; Unger, E. Nanotechnology 2001, 12(3), 238–244. (6) Brown, T. B.; Hancock, W. O. Nano Lett. 2002, 2(10), 1131–1135. (7) Prots, I.; Stracke, R.; Unger, E.; Bohm, K. Cell. Biol. Int. 2003, 27(3), 251–253. (8) Mitchison, T.; Kirschner, M. Nature 1984, 312, 232–237. (9) Schnackenberg, B.; Khodjakov, A.; Rieder, C.; Palazzo, R. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 9295–9300. (10) Stearns, T.; Kirschner, M. Cell 1994, 76, 623–637. (11) Faivre-Moskalenko, C.; Dogterom, M. Proc. Natl. Acad. Sci. U.S.A. 2002, 99(26), 16788–16793. (12) Holy, T.; Dogterom, M.; Yurke, B.; Leibler, S. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 6228–6231. (13) Urrutia, R.; McNiven, M.; Albanesi, J.; Murphy, D.; Kachar, B. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 6701–6705. (14) Nedelec, F.; Surrey, T.; Maggs, A.; Leibler, S. Nature 1997, 389, 305– 308. (15) Surrey, T.; Nedelec, F.; Leibler, S.; Karsenti, E. Science 2001, 292, 1167– 1171.
any polar organization of the MTs, the plus ends of the MTs radiated outward. When the kinesin constructs were used, the inherent directional movement of the N-terminal motors along assembling MTs ensured that these constructs were oriented with the minus end radiating outward. Although the use of C-terminal motor constructs could be used to reverse this polarity, such an approach would require an entirely new motor construct. In addition, these kinesin-construct approaches rely on the use of potentially sensitive, fragile, and short-lived biological motors for assembly. In the present work, we describe a new method for the assembly of polar-oriented synthetic microtubule organizing centers (POSMOCs). Such centers can be assembled such that MTs are oriented with their “plus” ends directed toward a central organizing particle. Alternatively, by simply modifying the order of component polymerization and assembly, POSMOCs may be created such that MT “minus” ends are directed centrally. As the methods used to prepare both “plus-in” or “minus-in” POSMOCs are similar, we will focus on the “plus-in” configuration for conciseness. The synthesis of these “plus-in” POSMOCs is achieved through a hierarchical assembly strategy, shown in Figure 1. First, the supramolecular polymerization of the MTs must be adapted to assemble hybrid MTs, where the “plus” and “minus” ends of an MT are chemically distinct from one-another. These distinctly functionalized segments would serve as chemical “handles” during assembly. Next, central particles must be created, such that particle size and chemical function must be tailored to the chemistry of the hybrid MTs. Finally, complete assembly is achieved using the chemical “handles” on the hybrid MTs to direct the selective attachment of the plus ends of the MTs to the core particles (or the minus ends for “minus-in” POSMOCs). The key to our approach for POSMOC assembly involves the ability to control the supramolecular assembly of MTs, creating segmented MTs in which one segment provides a functionalized handle for attaching the desired end of the MT to a central particle. This multistage process is shown in Figure 1. For the “plus-in” POSMOC synthesis, the hybrid MTs were synthesized by first polymerizing a relatively long segment of rhodamine-labeled tubulin in the presence of N-ethylmaleimide (NEM)-labeled tubulin, known to preferentially inhibit polymerization in the “minus” direction.16,17 This step ensured that the (16) Hyman, A.; Drechsel, D.; Kellogg, D.; Salser, S.; Sawin, K.; Steffin, P.; Wordeman, L.; Mitchison, T. Methods Enzymol. 1991, 196, 478–485.
Letters
Langmuir, Vol. 24, No. 14, 2008 7041
Figure 2. Fluorescence microscope image series showing polar-functionalized MT movement over an array of kinesin motor proteins. The MT is moving with the “minus” end in front, and is dragging a streptavidin-coated microsphere (indicated by arrows), bound to the biotinylated “plus” end of the MT. The image series spans 23 s, each image being 7-8 s apart.
addition of any tubulin during polymerization would be at the “plus” end of the MTs. In a second polymerization step, 10-fold diluted biotinylated tubulin was added to the growing rhodaminelabeled MTs. The dilution of the biotinylated tubulin was a key step in directing the segmented polymerization of the MTs. Fygenson et al.18 previously showed that the supramolecular polymerization of the tubulin is a concentration-dependent process. Incubated at 37 °C, 5 mg/mL tubulin concentrations will spontaneously nucleate and polymerize MTs, much like homogeneous crystal nucleation and growth. At 37 °C, however, there exists a concentration range between approximately 0.2 and 0.8 mg/mL in which tubulin polymerization is expected to occur, but MTs are nucleated primarily from existing nucleation sites on existing MT seeds, a process more closely resembling heterogeneous crystal growth. By diluting the biotinylated tubulin to 0.5 mg/mL, we have found that rather than polymerizing spontaneously, the biotinylated tubulin will preferentially add to existing rhodamine-labeled MTs. Furthermore, the NEM-capping of the rhodamine-MTs ensured that this new growth occurred almost exclusively at the “plus” end of the composite MTs. Without the NEM inhibitor, biotin would have been present on both ends of the hybrid MT, and both ends (not just the desired plus end) would have been functionalized for binding to the central particle. (see Supporting Information Figure 1) Composite MTs for the “minus-in” POSMOCs were made with similar methods, but the biotinylated tubulin was first polymerized at 5 mg/mL (in the presence of NEM), and the rhodamine-labeled tubulin was diluted and polymerized second. It should be noted that the length of either section can be controlled by varying the growth time. In this way, it is possible to tailor the relative length of the biotinylated binding segment or the fluorescent rhodaminelabeled segment to suit a particular assembly strategy. It was this careful manipulation of the polymerization process that led to the formation of polar segmented MTs. To ensure that polymerization processes produced the desired sequence of segments for controlling polar orientations on binding to the central particle, the MT sequences were interrogated using an “inverted motility assay.”19,20 In this assay, MTs were adsorbed onto substrates covered with an anchored array of inverted (17) Phelps, K.; Walker, R. Biochemistry 2000, 39, 3877–3885. (18) Fygenson, D.; Braun, E.; Libchaber, A. Phys. ReV. E 1994, 50(2), 1579– 1588. (19) Bachand, G.; Rivera, S.; Boal, A.; Gaudioso, J.; Liu, J.; Bunker, B. Nano Lett. 2004, 4(5), 817–821. (20) Howard, J.; Hunt, A.; Baek, S. Methods Cell Biol. 1993, 39, 137–147.
Figure 3. Top: A fluorescence image of “plus-in” POSMOC illustrating the segmented functionalization of the MT. A 2.34 µm silica sphere was decorated with streptavidin-coated quantum dots, indicated by the bright ring around the microsphere. The biotinylated “plus” end of a MT then bound to this functionalized microsphere. A subsequent treatment with streptavidin-labeled quantum dots has decorated only the biotinylated region of the MT (arrows). The relatively dimly fluorescent rhodaminelabeled segment (bracket) is visible extending away from the organizing center. Bottom: A contrast-inverted version of the image on the top more clearly shows the rhodamine MT (bracket) extending away from the organizing center.
7042 Langmuir, Vol. 24, No. 14, 2008
Figure 4. (a) POSMOCs with the “plus” end directed towards the microsphere. (b) POSMOCs with the “minus” end directed towards the microsphere. In both cases, fluorescent rhodamine-labeled MTs can be seen extending radially from the silica microspheres, functionalized with Oregon green-labeled streptavidin.
N-terminal kinesin motors. These ATP-activated motors align themselves with the polar direction of each MT such that they propel the MTs with the “minus” end leading and the “plus” end trailing. Biotinylated segments were tagged with fluorescent, streptavidin-coated microspheres. For the “plus-in” hybrid MTs, the microsphere-tagged, biotinylated segments were expected to appear near the trailing end of the moving MTs. Image series obtained using a fluorescence microscope (Figure 2) clearly show that “plus-in” MTs bound fluorescent, streptavidin-coated microspheres (100 nm diameter) and that binding only occurred in the trailing segment as expected, confirming the success of our synthesis protocol. Similar results were obtained with MTs of reversed polarity, but the functionalized components were seen at the leading end of the MTs. In initial experiments, composite MTs were assembled into POSMOCs, by introducing the hybrid MTs to a suspension of relatively large (2.34 µm) streptavidin-coated microspheres (the cores of the POSMOCs), functionalized using fluorescent streptavidin. This fluorescent streptavidin is responsible for the bright fluorescence emission from the microsphere surfaces in Figures 3–5. This surface fluorescence not only made the microspheres visible, but also confirmed the successful streptavidin functionalization of the microspheres. Once the polymerized hybrid MTs were reacted with these functionalized core microspheres, MT segment labeling was used to confirm that the desired biotinylated segment was attached to
Letters
the streptavidin-functionalized core. Using streptavidin-labeled quantum dots to decorate a particle-bound “plus-in” MT (Figure 3) one can see that the biotinylated segment not only wrapped partially around the surface of the core particle, but also extended a short distance into the solution. The biotin-free, rhodaminelabeled “minus” end of the MT extended outward from the attached biotin segment into the bulk solution (away from the larger particle). When high concentrations of MTs were attached to central organizing particles, they formed POSMOCs with the oriented MTs (all “plus-in” (Figure 4a) or all “minus-in” (Figure 4b)) radiating outward in three dimensions. In controls where either nonbiotinylated MTs were used or the silica spheres were not coated with streptavidin, these POSMOC asters did not form. The assembly of both the “plus” and “minus” POSMOCs demonstrates MT organization around relatively large particles. In these cases, the centralized particles remained relatively discrete from one-another and acted as the “dominant” organizing component in the system, directing the arrangement of the assembling MTs. When these large microspheres (∼2.3 µm) were replaced with smaller streptavidin-coated microspheres (100 nm), however, the organizational relationship between the particles and the MTs shifted. (Figure 5). The particles still served as central MT-organizing agents, but these microspheres were small enough that the MT could exert some influence over the organization of the microspheres; the microspheres were no longer the clearly dominant organizing component in the system. Figure 5 shows that, as the organizing particles were bound to the functional ends of the MTs, the MTs effectively drew the particles together binding them in a centralized aggregate. In a sort of organizational symbiosis, the particles served to organize the MTs, while the MTs in turn served to organize the particles. The organizational influence of the MTs was even more evident when the microspheres were replaced with streptavidin-labeled quantum dots. In this case, the MTs were in fact the dominant organizing component, as quantum dots were effectively templated along the biotinylated segments of the MTs. Once bound to the MTs, the quantum dots served to effectively functionalize the ends of the MTs with streptavidin. This effective streptavidin functionalization then allowed the MTs to bind to one-another through these quantum dot functional “linkers.” The resulting asters consisted of polar MTs whose “plus” (or minus) ends were tightly concentrated around a central point rich in biotinylated MT segments and templated streptavidin quantum dots. Unlike the 100 nm microsphere case, though, the quantum dots were not exclusively aggregated at the POSMOC core; quantum dots are visible extending outward, lined up linearly along the radiating MTs (Figure 5). This example illustrates most clearly how the cooperative organizational properties of this assembly strategy can produce not only the aster-like structure, but how the MT architecture can in turn be used to organize nanomaterials such as quantum dots, serving to both aggregate and template quantum dot assembly. Clearly the size of the streptavidin-functionalized particles has a significant influence on the relationship between the MTs and the particles during this assembly process. This approach to POSMOC synthesis not only facilitates the assembly of artificial organizing centers, but more generally demonstrates a robust and versatile method for arranging potentially technically valuable MTs in three dimensions with controlled polarity. By carefully controlling and modifying MT polymerization, we have shown that distinct chemical function can be assigned to the polar ends of these biosupramolecular structures. Taking advantage of this engineered chemical function
Letters
Langmuir, Vol. 24, No. 14, 2008 7043
Figure 5. (A) Polar MT organization directed by large (2.3 µm) microspheres. Multiple MTs are packed around each discrete microsphere. (B) Polar MT organization around aggregated 100 nm microspheres. Although MTs are organized around a centralized particle core, the MTs clearly affect microsphere organization into aggregates. (C) Polar MT organization using streptavidin quantum dots. Not only do quantum dots drive the assembly of MTs into an aster-like structure, but the MTs also serve as templates directing quantum dot organization within the structure.
in a second level of assembly, these complex structures can then be organized into asters, where the polarity and organization of the constituent components can be affected through interactions with various functionalized micro- and nanomaterials. This robust, straightforward methodology can even be adapted such that the polarity of the organized MTs can be reversed, by simply changing the order of the steps in the assembly process. In addition, by varying the size of the particles used to drive MT assembly, we have shown that the nature of the resulting aster-like structure may change as the organizational relationships between the MTs and the particles shift. In each of these varied cases, however, the application of the stable and selective, though noncovalent, biotin-streptavidin linkage21,22 proved to be very effective. These results suggest that this assembly approach could be adapted to a variety of alternative architectures where this linkage could be employed, including micropatterned surfaces and surfacepatterned nanostructures.23–25 The present work serves as an important demonstration of an assembly strategy capable of directing the polar organization of the MTs into three-dimensional constructs. This biologically mediated assembly is a critical first step in a new process for controlled assembly of nanomaterial structures. Future work may lead to new applications that depend (21) Moy, V.; Florin, E.-L.; Gaub, H. Science 1994, 266(5183), 257–259. (22) Zhang, Z.; Moy, V. Biophys. Chem. 2003, 104, 271–278.
on directional control of nanocargo transport within complex three-dimensional MT networks. Acknowledgment. The authors acknowledge Dr. Haiqing Liu at Sandia National Laboratories for insightful comments and discussion and Jennifer Pincus for assistance in functionalizing silica microspheres. This work was supported by the Division of Materials Sciences and Engineering in the Department of Energy Office of Basic Energy Sciences and Sandia’s Laboratory Directed Research and Development Office. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy’s National Nuclear Security Administration under Contract DEAC04-94AL85000. Supporting Information Available: Detailed experimental methods and one figure containing a fluorescence micrograph of an MT, biontinylated on both ends, bound to two functionalized microspheres. This material is available free of charge via the Internet at http://pubs.acs. org. LA800500C (23) Hyun, J.; Ahn, S.; Lee, W.; Chilkoti, A.; Zauscher, S. Nano Lett. 2002, 2(11), 1203–1207. (24) Michel, R.; Reviakine, I.; Sutherland, D.; Fokas, C.; Csucs, G.; Danuser, G.; Spencer, N.; Textor, M. Langmuir 2002, 18(22), 8580–8586. (25) Zhou, D.; Bruckbauer, A.; Ying, L.; Abell, C.; Klenerman, D. Nano Lett. 2003, 3(11), 1517–1520.
