Receptor-Based Differences in Human Aortic Smooth Muscle Cell Membrane Stiffness
Cells respond to mechanical stimuli with diverse molecular responses. The nature of the sensory mechanism involved in mechanotransduction is not known, but integrins may play an important role. The integrins are linked to both the cytoskeleton and extracellular matrix, suggesting that probing cells via integrins should yield different mechanical properties than probing cells via non–cytoskeleton-associated receptors. To test the hypothesis that the mechanical properties of a cell are dependent on the receptor on which the stress is applied, human aortic smooth muscle cells were plated, and magnetic beads, targeted either to the integrins via fibronectin or to the transferrin receptor by use of an IgG antibody, were attached to the cell surface. The resistance of the cell to deformation (“stiffness”) was estimated by oscillating the magnetic beads at 1 Hz by use of single-pole magnetic tweezers at 2 different magnitudes. The ratio of bead displacements at different magnitudes was used to explore the mechanical properties of the cells. Cells stressed via the integrins required ≈10-fold more force to obtain the same bead displacements as the cells stressed via the transferrin receptors. Cells stressed via integrins showed stiffening behavior as the force was increased, whereas this stiffening was significantly less for cells stressed via the transferrin receptor (P<0.001). Mechanical characteristics of vascular smooth muscle cells depend on the receptor by which the stress is applied, with integrin-based linkages demonstrating cell-stiffening behavior.
Mechanical stresses on cells lead to many important physiological responses, including gene activation,1–6 cell remodeling and migration,7–10 and change in extracellular matrix structure.11 The mechanisms by which a cell senses and responds to its mechanical environment are unclear, although some models have been proposed that depend on, for example, cytoskeletal linkages.12,13 However, a common theme is that the integrins are prime candidates for cellular mechanical sensors, because integrins link the extracellular matrix to the cytoskeleton.11,14–18
The response of cells to mechanical forces may depend on the receptor on which the stress is applied. An extension of this observation is that the mechanical properties of the cell may be different according to which receptor is being stressed. To test this hypothesis, it is necessary to have a technique for applying controlled stresses to specific receptors on a cell. Methods such as membrane stretching, fluid shear stress, and atomic force microscopy are not optimal because they do not control stresses or have the capacity to target specific receptors. Optical traps have been used with some success to probe cell membrane properties but lack the capacity to exert large forces (eg, those in the nanonewton range) and cannot control stresses without feedback.
A relatively recent technology is the development of magnetic tweezers.19,20 By controlling the current passing through an electromagnet, the force that is applied to magnetic beads can be set at a prespecified level. Specific cell receptors can be targeted for the application of mechanical stress by coating the magnetic beads with adhesion molecules or antibodies. Thus, by tracking how the displacements of the magnetic beads change, as a function of force level, one can obtain an estimate of the resistance of a cell to deformation as characterized by that membrane receptor, referred to as “stiffness” in the present study. This stiffness can then be compared among receptors to determine whether there is a difference in receptor response. For the present study, the receptors of interest were the matrix receptors (eg, integrins) and the transferrin receptors.
To test the hypothesis that cellular stiffness as a function of applied force differs for integrins and transferrin receptors, single-pole magnetic tweezers were constructed. The material for the tweezers was a CMI-C rod (Cold Metal Products Inc), a high-permeability paramagnetic metal with relatively high saturation. The tweezer pole was wrapped 550 times with a 24-gauge copper wire, turning the pole into an electromagnet. The current that passed through the wires was controlled with a computer and power supply by use of a custom-written program. The computer program also tracked the position of the magnetic beads at a sampling rate of 25 to 30 Hz with use of a charge-coupled device camera (CCD-100, DAGE-MTI).
The trap was set on the stage of a microscope (Axiovert S100TV, Zeiss) with the use of a 5-df micromanipulator (MX110R, SD Instruments) so that spatial adjustments could be made if necessary. Calibration was performed using a fluid of known viscosity (eg, polysiloxane or high sucrose solutions) and applying Stokes’ law for the low Reynolds number motion of the magnetic beads. The calibration results indicated that the range of force is from 10 pN to 10 nN, depending on the distance from the trap and the amount of current being passed through the tweezers.
Paramagnetic microspheres (4.5-μm diameter) were obtained from Dynal. The tosyl-activated surface function permitted covalent bonding of antibodies and proteins to the bead surface. The manufacturer’s instructions were followed in the coating of the beads. Briefly, the beads were washed in sodium phosphate buffer, pH 7.3, and then resuspended with 5 μg of protein per every 107 beads, at 37°C for 15 minutes, while they were slowly agitated. BSA was then added to make the final concentration of the solution 0.1% BSA. The beads were agitated slowly overnight, after which the beads were washed 3 times in PBS with 0.1% BSA and once with Tris buffer with 0.1% BSA, pH 8.5. The final concentration of the beads is the same as the original stock (4×108 beads/mL).
Immediately before use, the desired amount of beads was resuspended in 0.5 mL of PBS+1% BSA, mixed for 5 minutes, and then resuspended in medium. This bead-laden medium was then placed on the cells at 37°C for 1 hour to allow bead attachment, after which the experiment was performed.
Gaskets (Delta-T Culture Dish, 0.15 mm thick, Bioptechs) for temperature-controlled stages were used. The gaskets were coated with fibronectin at a concentration of 2 μg/mL of Hanks’ balanced salt solution (HBSS) at 4°C overnight. On the day of plating cells, the fibronectin solution was washed off, and the gaskets were rinsed twice with HBSS.
Primary human aortic smooth muscle cells from heart transplant donors were grown in Dulbecco’s modified Eagle’s medium+10% FCS supplemented with penicillin/streptomycin solution. Cell passages between 2 and 6 were used. The cells were detached with the use of trypsin after a rinse with calcium- and magnesium-free HBSS. The cells were then isolated by centrifugation and resuspended in Dulbecco’s modified Eagle’s medium+10% FCS. The cell density for plating was 16 000 cells/cm2, and the cells were allowed 18 to 24 hours to attach to the gaskets. The cells were near confluent at the time of experimentation.
Mechanical Stimulation Experiment
Cells that were plated and loaded with beads were placed on a Bioptechs stage, and the temperature was controlled to 37°C (Delta TC3, Bioptechs). The magnetic tweezers were lowered until they could be observed in the field of view with use of a ×10 objective (10× Fluar, Zeiss) and were just touching the bottom of the gasket (Figure 1). The magnetic tweezers were set to run at various controlling voltages at just <0.5 Hz. Because a full wave constitutes a positive and negative cycle, the effective frequency exerted on the beads is 1 Hz. Beads were tracked by using centroid tracking in the same program used to control the tweezers (Figure 2).
To obtain similar displacements of beads, a maximum of 0.2 V of driving voltage was used for the anti–transferrin receptor experiments, and a maximum of 2 V was used for the fibronectin experiments. The magnetic tweezers were activated at the highest intended setting (0.2 V for anti–transferrin receptor beads and 2 V for fibronectin-coated beads) and held constant for 20 seconds to remove loosely bound beads. Then, representative beads were selected, and the displacements on the beads were tracked as the oscillating force was applied for 5 full cycles at maximum voltage. The experiment was repeated at 75% of maximum voltage (for the anti–transferrin receptor–coated beads, this was 0.15 V, and for the fibronectin-coated beads, this was 1.5 V).
After the beads were tracked, the motion of the beads was split into 5 intervals (1 for each full cycle) and then smoothed by using 5-point averaging (2 points before and after the data point). The maximum and minimum were then determined for each of the 5 intervals, with the difference representing the maximum range of motion of the beads in that interval. Beads with differences too large (>1 bead diameter) or too small (no discernible waveform) were discarded, and the experiment was repeated until 5 data points were gathered. The 5 differences were then averaged and designated the characteristic displacement of the bead. Each bead has 2 characteristic displacements: 1 at maximum voltage and 1 at 75% maximum voltage. The ratio of the displacements (displacement ratio=displacement at 75% maximum voltage/displacement at maximum voltage) was taken for each cell. A 2-tailed t test was performed on the displacement ratios, with statistical significance set at P<0.05.
The differences in driving voltages needed to obtain similar displacements suggest that the integrin-bound beads required ≈10 times more force than did the transferrin receptor–bound beads. The displacement ratios of the cells were significantly different (P=0.044) (Figure 3). There was no significant difference in the actual amplitudes of motion at the higher voltage (force) level (P=0.53, with typical amplitude 0.140 μm) and no significant difference in the mean distance of the beads to the magnetic tip (P=0.98).
These results suggest that the transferrin receptor is not as well mechanically supported as the integrins. Because the integrins are linked to the cytoskeleton and presumably serve as a mediator for mechanical signaling, it is not surprising that they should appear stiff to the external environment. It is somewhat surprising that such a large force needs to be applied in smooth muscle cells to get a response from the beads. At 0.2 V, the fibronectin-coated beads did not move more than noise; in fact, this was true up to 1 V for most of the fibronectin-coated beads. Other cell types do not require as much force for similar displacements (data not shown).
The large stiffness of smooth muscle cells stressed via integrins could be a result of a much larger number of linkages between the cell and the fibronectin-coated beads compared with the cell and the anti–transferrin receptor beads. However, the difference in displacement ratios cannot be explained simply by increasing the number of linkages. The fact that the displacement ratio for integrin bound beads is high suggests that the membrane may be strain-hardened or, similarly, that the cell may be strain-softened for the anti–transferrin receptor beads. Although bead rolling can contribute to the differences in the motion of the bead, it does not detract significantly from the analysis, because the amplitudes of the bead displacements at the higher force level were the same. This means that the difference in the displacement ratios is somehow related to the mechanical response of the receptors, even if this relationship is indirect. It is also possible that the differences stem from different magnetic trap properties (eg, saturations or hysteresis) at the force levels used; however, some preliminary data indicate that the ratios of forces used are similar for both bead types.
Matrix receptors may play an exaggerated role in smooth muscle cells, inasmuch as these cells form the basis for the structural integrity of the major arteries. The increased force needed to obtain small displacements and the strain hardening effect via fibronectin-coated beads can be reconciled with the hypothesis that matrix receptors serve as mechanotransducers; because the smooth muscle cells characteristically experience much larger forces than many other cell types, their “mechanosensors” need to be buffered against relatively small forces and changes in force. On the other hand, we speculate that transferrin receptors may not serve as mechanoreceptors. Further studies may help to clarify the apparent differences in the receptor-based stimulus described in the present study.
This work is supported by National Aeronautics and Space Administration grant 99-HEDS-02/03-108.
- Received April 28, 2001.
- Revision received June 18, 2001.
- Accepted July 18, 2001.
Bao X, Lu C, Frangos JA. Temporal gradient in shear but not steady shear stress induces PDGF-A and MCP-1 expression in endothelial cells: role of NO, NF κB, and egr-1. Arterioscler Thromb Vasc Biol. 1999; 19: 996–1003.
Chen KD, Li YS, Kim M, Li S, Yuan S, Chien S, Shyy JY. Mechanotransduction in response to shear stress: roles of receptor tyrosine kinases, integrins, and Shc. J Biol Chem. 1999; 274: 18393–18400.
Davies PF, Remuzzi A, Gordon EJ, Dewey CF Jr, Gimbrone MA Jr. Turbulent fluid shear stress induces vascular endothelial cell turnover in vitro. Proc Natl Acad Sci U S A. 1986; 83: 2114–2117.
Mow VC GF, Tran-Son-Tay R, Hochmuth RM. Cell Mechanics and Cellular Engineering. New York, NY: Springer Verlag; 1994.
Davies PF. Flow-mediated endothelial mechanotransduction. Physiol Rev. 1995; 75: 519–560.
Lee RT, Berditchevski F, Cheng GC, Hemler ME. Integrin-mediated collagen matrix reorganization by cultured human vascular smooth muscle cells. Circ Res. 1995; 76: 209–214.
Ingber DE. Cellular basis of mechanotransduction. Biol Bull. 1998; 194: 323–327.
Boudreau NJ, Jones PL. Extracellular matrix and integrin signalling: the shape of things to come. Biochem J. 1999; 339: 481–488.
Giancotti FG, Ruoslahti E. Integrin signaling. Science. 1999; 285: 1028–1032.
Luscinskas FW, Lawler J. Integrins as dynamic regulators of vascular function. FASEB J. 1994; 8: 929–938.