Undercuts enable nanoshuttle steering of microtubules

Research conducted at the University of Washington (Seattle, WA) and Sandia National Laboratories (Albuquerque, NM) demonstrates that carefully designed microstructures created photolithographically in conventional photoresists can make better "roads" for microtubules and other particles. Though the idea of these pathways is not new, the new configuration is both easier to implement and better at keeping particles trapped in the desired path—both without reducing the speed at which the particles travel. Further, the new channels are narrower than existing pathways, making them more like some of the biological structures they are emulating.

Nanotransportation is based on the idea of having tiny shuttles driven by biomolecular motors. The motor proteins come in two main types—kinesins and myosins—and are what make spermatozoa into swimmers and give muscles their pull. Their strength comes, in part, from their efficiency. Motor proteins are fueled by adenosine triphosphate (ATP) via a process that turns chemical energy directly into mechanical movement without losing it to heat. The proteins themselves look a little like a pair of legs with a torso (see Laser Focus World, August 2003, p. 13). When in active mode, they "walk" using a chemical-mechanical process that involves bonding with one chemical "leg," a leverlike movement that assists in releasing the other leg, and then bonding with the released leg. The ATP is critical to the bonding and releasing action.

Conveyor-belt system

The idea of passive nanotransportation is based on using biomolecular motors (motor proteins) as the moving part of a conveyor-belt-type system. Instead of having them doing the walking, the legs push particles around—specifically, they push microtubules that have the appropriate chemical structure to allow the walking action. Linkers can be added to the shuttles to enable them to carry cargo of various types; because gravity is unimportant at this scale, the orientation of the tubules, cargo, and motor proteins are irrelevant.

One problematic area has been steering. Simply patterning a surface with roads of kinesin will work to a limited extent, but problems occur when a microtubule wanders to the edge of the track. Because there is no mechanism for it to reorient itself, the tubule can simply get stuck. One attempt to improve on this involved building sidewalls, also covered in motor protein. Although this approach was partially successful, the shuttles could still "jump the track" by climbing up the sidewalls. Yet another approach involved having nonfouling (inhospitable) sidewalls; this has been very successful at guiding the tubules. But because the additional nonfouling surface must be deposited, the whole system then becomes more expensive and difficult to fabricate.

The new technique involves using a reverse photoresist to create undercut channels. After an initial exposure of the resist—which has been spun onto a glass substrate—the substrate is baked and the exposed areas crosslink. An aggressive final exposure exposes the areas that have not been crosslinked, and the resist is developed and treated with oxygen plasma to oxidize it and clean the class. According to Henry Hess of the University of Washington, the resulting undercut is caused by underexposed areas being more soluble in developing fluid (see figure). Despite the substrate being transparent, researchers used an exposure time intended for reflective semiconductor surfaces, giving the deeper material its softness and making it easy to remove.

By creating an undercut surface and coating it with motor proteins, researchers have created a structure that can trap and channel microtubules used as shuttles in a nanoconveyor-belt system.

All that is necessary once the structure is finished is to coat it with the motor protein; antifouling materials are unnecessary because the microtubules, once they enter, are trapped in the undercut region. In an experiment, not a single tubule of the 43 they imaged was able to climb up to the bottom surface.¹ On the contrary, says Hess, most of those that started on the top surface eventually descended. Further, the new technique allows the channeling of particles in partially closed submicron structures as opposed to the open, micron-wide structures demonstrated previously. This, says Hess, makes them more similar to biological structures such as axons.

REFERENCE

1. Henry Hess et al., Nano Letters 3(12), 1651 (2003).