© 2003 Robert A. Freitas Jr. All Rights Reserved.

Robert A. Freitas Jr., Nanomedicine, Volume IIA: Biocompatibility, Landes Bioscience, Georgetown, TX, 2003

15.5.4.1 Force Threshold for Biological Response

Mechanical stresses modulate cell function by either activating or tuning signal mechanotransduction pathways, via various connections between the internal cytoskeleton, the ECM, and traditional signal transducing molecules [5332]. Vibrations or tugging forces applied to the ECM are transmitted to focal adhesions, attachment points at the plasma membrane surfaces of nearby tissue cells. ECM-cytoskeletal couplings occur through transmembrane integrins having greater mechanical stiffness at high applied stress (>1 N/m²) than similar couplings through transmembrane E-cadherins [3966]. At the tissue cell surface, gated transmembrane channels are activated by simply stretching the plasma membrane, or by tension or stress development in cytoskeletal elements associated with the cell membrane [3967, 3968]. Mechanical strain deformation modulates the morphology, metabolism and activation of: chondrocytes in articular cartilage [3790, 3969]; airway [3970] and vascular [3794] smooth muscle cells; primary astrocytes and glioma cells [3971], neurons [3789], and auditory sensory cells [3972]; endothelial cells [3809]; and even prostate cancer cells [3973]. Simple fluid nozzling or mechanical prodding elicits mechanosensitivity in rat myocytes [3858] and neurons [3789, 3974]. Mechanical pulling force applied by micropipette to the integrin-containing dot-like focal adhesion complexes, ~1 micron in diameter, between cell and ECM leads to local assembly and elongation of these structures into streak-like focal contacts (3-10 microns long). Focal complexes thus serve as cellular mechanosensors exhibiting directional assembly in response to locally applied force [3975]. The response to mechanical strain can take place over a period of years [3976] or can occur as fast as minutes [3788, 3858] or even seconds [3971, 3977].

Biochemical transduction of mechanical strain has been investigated quantitatively in bone cells during normal loading. Linear strains of <0.05% are nonstimulative; those between 0.05-0.15% maintain normal bone mass; strains >0.15% stimulate osteoblasts to increase bone mass [3978-3980]; and linear strains >1% induce osteoblasts to alter morphology, becoming fibroblast-like [3981]. For instance, chick osteoblasts subjected to 1.3% applied strain as a 0.25 Hz dynamic spatially uniform biaxial strain for 2 hours experienced elevated osteopontin expression, rising to a maximum 4-fold increase after 9 hours from the beginning of strain onset [3982]. Differentiated mechanosensitive mouse podocytes (glomerular cells) cultured on silicone membrane and subjected to a 0.5 Hz biaxial cyclic stress for up to 3 days at 5% linear strain experienced a reduction in cell body size, a thinning and elongation of cellular processes, and a reversible reorganization of the actin cytoskeleton, uniquely involving formation of radial stress fibers with the disappearance of transverse stress fibers [3793]. A 15% surface-to-surface strain imposed on articular cartilage ECM triggered a shrinkage of chondrocyte cell height (-14.7%) and volume (-11.4%) and a shrinkage of chondrocytes nucleus height (-8.8%) and volume (-9.8%) as well [3983].

What is the force threshold for biological response by tissue cells to the mechanical disruption of cell-ECM contact? There is a range of forces* from 0.08-400 pN that produces a span of responses in different cells, given that mechanosensitivity is a near-ubiquitous property of cells [3858]. Some biological response may occur near the lower end of this range of forces. For instance, when laser tweezers applied a force of 7 pN to individual bone- and cartilage-derived cells in vitro, an immediate increase in intracellular Ca⁺⁺ ions was observed in human-derived osteoblasts, and force applied to different regions of a cell produced a variable response [3993]. (The response is inhibited by the calcium channel blocker nifedipine [3993].) In rat femur-derived osteoblasts, the Ca⁺⁺ elevation in response to a similar load was lower, and was entirely absent in primary chondrocytes and the osteocytic cell line [3993]. Osteoblasts also express certain proteins when subjected to fluid shear stress of ~1.2 pN/micron² (inhibited by intracellular calcium chelator or by the calcium de-storage agent thapsigargin) [3994]. Mechanical shear stress sufficient to generate acute release of prostaglandin E2 (PGE2) in isolated chicken osteocytes (bone mechanosensory cells that translate mechanical signals into biochemical bone metabolism-regulating stimuli necessary for the adaptive process) was induced by 10 minutes of 5 Hz pulsating fluid flow at 0.7 pN/micron² [3995]. Again, several inhibitors are reported [3995]. Optical tweezer manipulation of neutrophils with ~pN forces does not damage the plasma membrane but instead stimulates a mechanically-inducible, membrane channel-mediated influx of extracellular Ca⁺⁺ into the cell [3791].

* We can get a sense of the magnitudes involved by reviewing some of the forces required to mechanically separate, or to extract, integral proteins or ECM-attachment molecules from the cellular plasma membrane. For example:

(1) 0.08-0.35 pN/bond to separate CD2 molecules (expressed on Jurkat cell surface) from two isoforms of transmembrane or glycophosphatidyl LFA-3 (lymphocyte function-associated antigen 3) that are already incorporated into lipid bilayers [3984, 4647];

(2) 2.1 pN to separate platelet cell surface-activated GpIIb-IIIa integrin from an attached fibrinogen molecule, compared to 0.57 pN for nonactivated integrin [3984];

(3) 4 pN/bond to separate a human T cell and its target cell doublet, with 400 bonds/micron² [3984, 3990, 3999];

(4) 2.8-11 pN/bond to separate murine fibroblast cells bearing ICAM-1 (intercellular adhesion molecule 1) and murine T cells expressing LFA-1 (lymphocyte function-associated antigen 1) [3984, 4648];

(5) 6-10 pN to separate each connexin-32 hepatic cell gap junction unit [3985];

(6) 10-20 pN to extract glycophorin A from RBC membrane [3986];

(7) 25-45 pN for L-selectin (CD62L) extraction from neutrophil membrane in 1-2 sec [3987];

(8) 35-85 pN for CD45 extraction from neutrophil membrane in 1-11 sec [3987];

(9) ~60 pN to extract PSGL-1 selectin from platelet membrane [3988];

(10) 65 pN to rupture bond between lectin and RBC membrane-bound glycolipids [3989];

(11) 60-130 pN to extract beta2-integrins (CD18) from neutrophil membrane in 1-4 sec [3987];

(12) ~100 pN to extract integral glycoprotein from cell lipid bilayer (RBC membrane) [3990];

(13) 165 pN to rupture P-selectin bond with leukocyte-membrane-bound P-selectin glycoprotein ligand-1 [3991]; and

(14) 40-400 pN to separate a pair of cell adhesion proteoglycan molecules on marine sponge cell surfaces [3992].

Nevertheless, somewhat higher forces from nanorobotic histonatation (Section 9.4.4) and cytopenetration (Section 9.4.5) might be tolerated without reaction by the ECM because such higher forces are frequently applied by individual motile cells traversing human tissues. For example, the adhesion strength for the protozoan Amoeba proteus has been measured as ~100-1000 nN [3996], giving a transient adhesion force of 100-1000 pN/micron² over a focal contact area of ~1000 micron² [3997]. The tension force exerted by a single fibroblast during locomotion has been measured as ~165 nN [3998], or ~1000 pN/micron² (1000 N/m²). Cell-cell adhesion of T cells and target cells is ~1500 pN/micron² [3999], and live cells may apply ~5500 pN/micron² constant stress at focal adhesions to ECM [3977]. The foregoing would suggest conservative thresholds for biological response from human tissue cells subjected to nanorobotic mechanical operations of perhaps ~10 pN absolute force, ~1-100 pN/micron² shear stress, or ~0.1% linear strain on cellular membrane. If these response thresholds unavoidably must be exceeded, many inhibitors of cellular mechanosensitivity are known [3787-3794] but these must be delivered at µM-mM concentrations to the immediate vicinity of the cell in order to be effective.

Recent experiments with Lymnaea neurons [4000] found both the expected result that the probability of mechanosensitive channels being open is proportional to membrane tension, and also the unexpected result that many channels appear insensitive to mechanical stimuli in situ. Failure to elicit mechanocurrents from in situ cells having abundant channels suggests that channels may normally be protected from mechanical stimuli in situ, and that only traumatized cell membrane (i.e., traumatized cortical cytoskeleton) may be unable to prevent transmission of mechanical stimuli to plasma membrane channels, a theory the authors call “mechanoprotection” or “capricious mechanosensitivity”. (It is already known that cultured chondrocytes must be externally loaded at >0.5 MPa to prevent disassembly of the vimentin components of their cytoskeleton [4001].) If these results are confirmed for human tissue cells, then in situ tissue cells would be suppressed from reacting to nanorobots whose passage produces only nondestructive mechanical strains at their surfaces. Additional research is clearly required.

Similarly, Zhang and Hamill [5662] found that mechanical stimulation of Xenopus oocytes by inflation, aspiration, or local indentation (even to the point of membrane damage) fails to activate mechanosensitive ion channels, which they attribute to changes in membrane geometry (e.g., buffer membrane drawn from surface microvilli ~1.4 µm in length, 0.12 µm in diameter, and numbering 6-7/µm² on the surface [5663]). The discrepancy between patch and whole cell mechanosensitivity (i.e., “mechanoprotection”) arises because animal cells have an excess membrane area (compared to the minimum necessary to cover their volume if enclosed as a sphere) that tends to buffer changes in bilayer tension caused by mechanical stimulation. Notes Owen Hamill [personal communication, 2002]: “For the specific case of a nanorobot, a problem could arise if the robot applied a local increase in tension in the bilayer beyond its elastic limit and thereby ruptured the patch. I would presume that nanorobots would be less able to apply global changes in cell tension. Local changes could occur, for example, if the robot legs in sticking to exposed area of the bilayer (after the external matrix had been removed) stretched a patch of bilayer similar to that what occurs when suction is applied to a patch clamp pipette.”

Last updated on 30 April 2004