© 1999 Robert A. Freitas Jr. All Rights Reserved.

Robert A. Freitas Jr., Nanomedicine, Volume I: Basic Capabilities, Landes Bioscience, Georgetown, TX, 1999

9.4.7.4 Steering and Control

A complete understanding of motile cell mobility biological control systems is not yet available. However, present knowledge¹⁵⁶⁴ is sufficient to support the conclusion that virtually complete command of cellular mobility systems by in cyto medical nanorobots should be feasible. Such control may be effected by either biochemical or mechanical means.

The "amoeboid" movement (Section 9.4.3.7) of leukocytes and fibroblasts relies primarily upon temporary extensions of the cell surface rather than major movements of cytoplasm. These temporary extensions, such as the small filopodia and the larger lamellipodia, are surface evaginations involving bundles of actin filaments or microtubules* in parallel array, often comprising a large fraction of the total surface area of the cell. Motion is made possible by an internal system of bracing which is built or modified according to the needs of the cell. Immediately below the plasma membrane is a region of densely packed and interconnected actin microfilaments called the cell cortex (Section 8.5.3.11). A leukocyte moves by breaking down actin filaments and rebuilding them in the proper orientation, pushing the cell in a new direction and allowing the posterior regions to retract.³¹²

* Treadmilling and dynamic instability control microtubule length in vivo.¹⁵⁶⁰

Fiber bundles isolated by microlaser surgery from a glycerinated fibroblast have been shown to contract with the addition of ATP.⁹³⁸ Focal attachments between substrate and cell are modulated both by external chemotactic gradients on the substrate and by internal biochemical regulatory pathways (second messenger molecules; Section 7.4.5.1); similar signal transduction pathways control the polymerization and disassembly of filaments and microtubules comprising the podial extensions. Regulatory proteins that bind actin monomers and filaments, some of which function in response to transmembrane signaling, include ADF, adseverin, caldesmon, cap Z, cap 100, cofilin, gelsolin, MCP, phospholipase C-g, profilin, and severin.¹⁵⁶⁴ Manipulation of the internal concentrations and spatial distributions of appropriate intracellular messenger molecules and ATP should permit in cyto medical nanorobots to command these biochemical pathways and to regulate localized polymerization processes,¹⁹⁴³ thus controlling the speed and direction of motile cell locomotion.

As another example of an intracellular mediator system that may be subject to direct nanorobotic control, consider the Ras superfamily of small guanosine triphosphatases (GTPases). One member of this superfamily, known as Rho, can be activated by the addition of extracellular ligands (such as lysophosphatidic acid), leading to the assembly of contractile actinmyosin filaments (stress fibers) and associated focal adhesion complexes.¹⁵³⁸ Rho acts as a molecular switch to control a signal transduction pathway linking plasma membrane receptors to the cytoskeleton.¹⁵³⁹ Rac, another member of the Rho subfamily, can be activated by a distinct set of agonists such as platelet-derived growth factor (PDGF)* or insulin, leading to the assembly of a meshwork of actin filaments at the cell periphery to produce lamellipodia and plasma membrane ruffles.¹⁵⁴⁰ Activation of Cdc42, another Rho subfamily member, induces actin-rich filopodial surface protrusions that are also associated with distinct integrin-based adhesion complexes.^1541,1542 There is significant cross-talk between GTPases of the Ras and Rho subfamilies -- Ras can activate Rac (thus Ras induces lamellipodia), Cdc42 activates Rac (thus associating filopodial and lamellipodial formation), and Rac can activate Rho.^1540,1541 Members of the Rho GTPase family appear to be key regulatory molecules linking surface receptors to the organization of the actin cytoskeleton.¹⁵³⁹ By 1998, about 10 GTPase-activating proteins and 3 guanine nucleotide dissociation inhibitors (both potential down-regulators of GTPase activity), and at least 20 cellular effector targets with relevant protein-protein interaction domains, had been described.¹⁵³⁹ Local cytoplasmic injections or extractions of nanomolar concentrations of these second-messenger substances should permit full control of cytovehicular direction and speed.

* PDGF concentration in human serum is 50 ng/cm³ (Appendix B); optimum concentration is 1-20 ng/cm³,³¹⁵³ or ~1 molecule/micron³, for chemotactic response of neutrophils and monocytes during wound repair (Chapter 24). Fibroblasts need the higher concentration in this range (~20 ng/cm³); PDGF stimulates a full mitogenic response.¹⁵⁵⁴

Less subtle biochemical methods of cytoskeletal manipulation are also available to nanorobotic pilots. For example, the drugs colchicine and taxol disrupt microtubule function in distinctly different ways. Colchicine binds to tubulin, strongly inhibiting its further assembly into microtubules and fostering the disassembly of existing microtubules;⁹³⁹ two other alkaloids, vinblastine and vincristine, also reversibly inhibit microtubule assembly.⁹³⁶ (Microtubules are always losing and reacquiring tubulin dimers, and these three drugs prevent reassembly, so eventually the tubules would disappear entirely.) Colchicine diminishes neutrophil migration into inflammatory foci.⁹³⁶ By contrast, taxol binds tightly to microtubules, stabilizing them and causing much of the free tubulin in the cell to assemble into microtubules.⁹³⁹ Similarly, the drug phalloidin (e.g., ~5 nanomoles/cm³)¹⁵⁵⁷ blocks the depolymerization of actin, thereby stabilizing microfilaments, whereas cytochalasin B inhibits the polymerization of actin microfilaments.⁹³⁹ Drug molecule diffusion time over ~1 micron ranges is ~1 millisec, consistent with ~KHz cycling. S. Smith notes that the use of oncogene products like members of the Ras supergene family, coupled with cytotoxins and genotoxins like phalloidin and vinblastine for steering, could transform a leukocyte. The risk of generating a leukemia must be avoided, perhaps by implanting a timed self-destruct signal in the cytovehicle before initiating cytocarriage.

Cytovehicles may also be piloted by palpating the cytoskeleton mechanically -- for instance, by applying tension to selected parts of the cytoskeleton using nanorobot manipulators. ECM-mediated mechanical forces can pull cells into square and rectangular shapes with sharp 90° corners.^718,1555 Mechanical manipulation may also be needed to drive cells into places they will not normally go. For instance, cells parked on experimentally-created micropatterned adhesive islands surrounded by nonadhesive (PEG-coated) boundaries cannot be induced to step off the islands even when stimulated with high concentrations of soluble growth factors.¹⁵⁵⁵ Mechanical signals are transduced into biochemical signals via mechanosensory plasma membrane channels¹⁵⁰⁶ and by other means.

How fast can cytovehicles be driven? Even using supranormal concentrations of biochemical control molecules combined with close placement to active sites in order to minimize diffusion time delays, a purely biochemically-mediated control system is probably limited to the maximum speeds of actin polymerization or osmotic inflation, or ~1-10 microns/sec. However, purely mechanical cytovehicular control systems can be cycled at much higher speeds, at least in tension. Individual actin microfilaments will tolerate a maximum tensile stress of ~108 pN (Section 8.5.3.11); the binding force between cytoskeleton and cell cortex at focal adhesions is of similar magnitude. Let us assume that the piloting nanorobot manipulates just a single fiber to drive each ~1 micron diameter lamellipodium, and that this single fiber is not driven beyond 10% of its theoretical failure strength, a very conservative limit. Using Eqn. 9.73 to crudely approximate the viscous drag force on the protuberance as it moves through interstitial fluid, each lamellipodium can be flexed at speeds up to ~500 micron/sec. Manipulating more fibers adds to the margin of safety. (Cyclosis in mature plant cells such as Nitella and Chara produces circular motions of some cell components as fast as 100 microns/sec.⁹³⁸) Similar manipulation of the much stronger intermediate fiber (IF) cytoskeleton (Section 8.5.3.11) would in theory permit speeds up to ~0.1 m/sec, but such aggressive forces might rip the IF fibers loose from their intracellular moorings. Another potential concern is that rapid cycling will stimulate unwanted peripheral actin polymerization or depolymerization reactions unless they are actively suppressed.

Electric fields may also drive leukocyte motion (Section 4.9.3.1). For example, a mild electric current induces lymphocytes to travel in the same direction of the current ("electrotaxis") at speeds up to ~0.3 microns/sec.⁸⁴⁸ Small electric gradients have been shown experimentally to stimulate leukocyte diapedesis (Section 4.9.3.1),⁶⁹⁰ though multiple nanorobots would probably be required to generate the necessary currents.

Last updated on 22 February 2003