Polarity orientation of microtubules and its applications with motor proteins

Method 1: The microfluidic channel used for Method 1 shown in figure 1 was fabricated by a standard PDMS molding process and bonding to a coverslip. The channel was 30 mm in length, 30–500 μ^m in width and 80 μ^m in height. Thick negative photoresist (SU-8 2050, Microchem) was patterned on a silicon wafer as a mother mold for the channels. PDMS prepolymer (Sylgard 184, Dow Corning) was then poured on the mold and cured. Multiple channels were transferred onto the PDMS surface. The replica was placed on a coverslip and pressed to remove bubbles without any adhesive. The adhesion between the coverslip and the PDMS is sufficient to manipulate buffers without leakage. The inlet and outlet were prepared by a punch to introduce assay buffer solutions.

Zoom In Zoom Out Reset image size

Method 2: Submicrometer scale channels, ranging from 500 nm to 2 μ^m in width, on the PDMS replica were fabricated to confine a single MT filament. The channels were sealed with a glass coverslip. Parallel channels connect a pair of micro-scale access channels. These were necessary to inject liquids to the submicrometer channels by pipette manipulation. Two different mask layers were patterned for the mother mold fabrication on a silicon wafer. As the first layer, a negative electron beam (EB) resist (Microposit SAL601-SR7, Shipley) for submicrometer channels was spin-coated on a silicon wafer (30 s at 4000 rpm), and exposed (15 μ^{C  cm −2}, 20 keV) by an EB-patterning SEM (Tokyo Technology). Second, an UV resist (Microposit S1805, Shipley) was spin-coated onto the patterned area (30 s at 4000 rpm) and access micro-scale channels were defined. The silicon wafer was etched down by DRIE (Multiplex-ICPRIE, STS) using the patterned EB and UV resist. The channel structure illustrated in figure 2(A) was transferred to a PDMS slab by the same process mentioned for Method 1.

Zoom In Zoom Out Reset image size

Method 3: Nano tweezers, as shown in figure 3(A), were initially made for DNA manipulation [6]. The tweezers were fabricated using an SOI wafer. Fabrication was based on reactive ion etching, local oxidation and anisotropic etching of silicon, as explained in detail elsewhere [7]. First, a thin ^Si ₃ ^N ₄ layer was deposited by low-pressure chemical vapor deposition (LPCVD) and patterned. Then, ^Si ₃ ^N ₄ and the Si over layer were etched by reactive ion etching (RIE). A local oxidation of silicon process was used to grow ^SiO ₂ on the top and sidewalls of the uncovered Si surface. The ^Si ₃ ^N ₄ layer was then removed and a KOH wet anisotropic etching of Si from the exposed area was performed to obtain {111} facets. The intersection of three planes, i.e. the bottom 〈001〉 plane protected by the buried oxide, the vertical 〈010〉 plane protected by LOCOS oxide and the 〈111〉 plane obtained by KOH etching, made sharp opposing tips. As the last step, the buried oxide was removed using vapor HF, and the handling Si was structured due to deep reactive ion etching.

Zoom In Zoom Out Reset image size

Method 1: The orientation process was performed in the PDMS channel. Kinesin in BRB80 buffer solution was incubated for 10 min to coat the glass surface and the excess kinesin was washed out by BRB80. The MTs were trapped in parallel to the buffer flow on the kinesin-coated surface, as shown in figure 1(a). Although MTs are physically aligned by the fluid force, these MTs have random polarity with respect to the functional plus/minus orientation. The continuous buffer flow with 0.5–1 mM ATP was utilized to orient the immobilized MTs. The continuous flow, however, removes MTs gliding upstream, while almost all of the MTs gliding downstream remained on the kinesin-coated glass surface. Figure 1(b) shows some MTs removed and rectified by the flow. BRB80, including 0.1% glutaraldehyde, was introduced to immobilize the MTs. This resulted in the immobilization of gliding MTs with physical and functional orientation (figure 1(c)). 0.1 M glycine in BRB80 was also injected for 10 min to block aldehyde groups.

Method 2: Figure 2(A) illustrates the detailed experimental procedure to orient and immobilize a MT in each channel, and to transport kinesin-coated beads on it. Kinesin was first introduced to drive MTs on the glass surface. Excess kinesin was washed out by BRB80 buffer solution, and MTs were injected into area A in figure 2(A)(1). They were trapped by kinesin on the channel surface, but did not enter the submicrometer scale channels, because the buffer flow was only applied to area A with area B closed. Unbound MTs in area A were removed by buffer flushing, which allows us to have trapped MTs. ATP injection causes the gliding assay of MTs and some of them glide into the submicrometer channels (figure 2(A)(2)). The number of MTs in each channel depends on the concentration of MTs, the size of the channels and the assay time. The number of MTs introduced into a channel increases with an increase in these parameters. Once the movement of the MT head was confined at the entrance of a submicrometer channel, it kept gliding into the channel without making a U-turn (figure 2(A)(3)). This is caused by the stiff structure of MTs compared with other filaments, such as actin and DNA.

Method 3: In this method, MTs were individually manipulated by the fabricated tweezers shown in figure 3(A). Pick-and-place assembly of individual MTs can be divided into three steps: isolation, capturing and relocation. Isolation is the step where MTs are bridged between the parallel walls (figure 3(B)). Using polarity marked microtubules (PMMT), whose minus end is labeled brighter than the plus end by fluorescent dye, we can easily identify MT polarities. In the capturing step, micro tweezers are used to approach a PMMT bridged between Cytop walls, as shown in figures 3(C) and (D). Physical contact between the PMMT and tweezers tips results in bridging the MT between the tips. After capturing the MT, relocation is performed by carrying the MT to a Poly-L-lysine coated surface, on which MTs are much more adhesive than with tweezers tips. Repeating the process enables multiple MT relocations, as shown in figures 3(E)–(G) [3].

We utilized kinesin-coated beads to evaluate MT polarities oriented by the above three methods. Since conventional kinesin moves to the plus end of a MT, we are able to identify each MT polarity by observing bead motion under an inverted microscope (Olympus IX71). Carboxylated polystyrene beads with a diameter of 0.33 μ^m were nonspecifically coated with full-length conventional kinesin with GST-tag. Beads were injected on a MT fixed surface by each polarity orientation method. Then, beads started to move on the 1 mM ATP injection. We evaluated the ratio of the moving beads moving in the expected and unexpected directions for Method 1. For the other two methods, we focused a few countable numbers of beads to examine the MT polarities.