Remanent cell traction force in renal vascular smooth muscle cells induced by integrin-mediated mechanotransduction

Abstract

It was previously demonstrated in isolated renal vascular smooth muscle cells (VSMCs) that integrin-mediated mechanotransduction triggers intracellular Ca²⁺ mobilization, which is the hallmark of myogenic response in VSMCs. To test directly whether integrin-mediated mechanotransduction results in the myogenic response-like behavior in renal VSMCs, cell traction force microscopy was used to monitor cell traction force when the cells were pulled with fibronectin-coated or low density lipoprotein (LDL)-coated paramagnetic beads. LDL-coated beads were used as a control for nonintegrin-mediated mechanotransduction. Pulling with LDL-coated beads increased the cell traction force by 61 ± 12% (9 cells), which returned to the prepull level after the pulling process was terminated. Pulling with noncoated beads had a minimal increase in the cell traction force (12 ± 9%, 8 cells). Pulling with fibronectin-coated beads increased the cell traction force by 56 ± 20% (7 cells). However, the cell traction force was still elevated by 23 ± 14% after the pulling process was terminated. This behavior is analogous to the changes of vascular resistance in pressure-induced myogenic response, in which vascular resistance remains elevated after myogenic constriction. Fibronectin is a native ligand for α₅β₁-integrins in VSMCs. Similar remanent cell traction force was found when cells were pulled with beads coated with β₁-integrin antibody (Ha2/5). Activation of β₁-integrin with soluble antibody also triggered variations of cell traction force and Ca²⁺ mobilization, which were abolished by the Src inhibitor. In conclusion, mechanical force transduced by α₅β₁-integrins triggered a myogenic response-like behavior in isolated renal VSMCs.

myogenic vasoconstriction is an autoregulatory mechanism for adjusting local vascular resistance to the increase in arterial pressure, which fluctuates over a wide range under different physiological and pathophysiological conditions (27, 43). In the renal circulation, pressure-induced myogenic constriction in the afferent arteriole is an important protective mechanism to prevent the transmission of elevated systemic pressure to the glomerular capillaries, a critical determinant in the progression of glomerulosclerosis in diabetes, hypertension, and end-stage renal disease. However, the mechanisms that transduce the changes in perfusion pressure into vascular smooth muscle cells (VSMCs) to initiate myogenic constriction are still elusive (7, 15, 17, 28). Many studies on mechanotransduction of myogenic response have focused on the mechanosensitive gating of stretch-sensitive channels in triggering membrane depolarization and Ca²⁺ influx. However, stretch-sensitive ion channels cannot provide a sustained error signal to maintain the myogenic constriction when the stretch vanishes as the vessel contracts (9, 10, 42). In contrast, integrins are known to transduce extracellular mechanical force into intracellular biochemical signals to affect VSMC contractility (1, 20, 21, 32). We have previously shown that application of mechanical force to renal VSMCs with paramagnetic beads coated with integrin ligands activates subcellular Ca²⁺ release from ryanodine receptors in the form of Ca²⁺ sparks (1). The coupling of Ca²⁺ sparks to the Ca²⁺-activated Cl⁻ channel may account for the membrane depolarization in VSMCs required for contraction (22). However, it has not been determined whether the integrin-mediated mechanotransduction is sufficient to elicit VSMC contraction for the myogenic response. Cells are anchored to the substratum via focal adhesions, which are composed of cytoskeleton, integrins ,and extracellular matrix. VSMCs exert cell traction force through focal adhesions (34). In the present study, cell traction-force microscopy was used to monitor the changes in cell traction exerted by renal VSMCs on a flexible substrate to examine integrin-mediated contraction at the single cell level.

MATERIALS AND METHODS

Isolation and culture of renal VSMCs.

All experiments were performed under protocols approved by the University of South Florida's Animal Care and Use Committee. Male Sprague-Dawley rats from Harlan Laboratories (120–200 g) were euthanized by anesthetic overdose (5% isoflurane in a chamber through a vaporizer). Renal vascular smooth muscle cells were isolated enzymatically from dissected arcuate arteries and cortical radial arteries, which are capable of responding to changes in perfusion pressure (3), and the yield of isolated cells from these arterioles was much higher than the afferent arterioles (14). The arteries were first digested with 0.15% papain (Sigma) for 6 min and then digested with 0.2% collagenase type 4 (Worthington), 0.1% trypsin inhibitor (Sigma), and 0.05% elastase (Sigma) for 4–6 min.

All digestions were performed in low calcium dissociation solution containing the following (in mM): 119 NaCl, 4.7 KCl, 1.2 MgSO₄, 1.18 KH₂PO₄, 24 NaHCO₃, 5.5 glucose, 10 HEPES, and 50 μM CaCl₂, pH 7.4 at 37°C. The vessels were then triturated, and the freshly isolated VSMCs were plated onto glass-bottomed Petri dishes coated with Matrigel (BD Biosciences). Some VSMCs were resuspended in DMEM with 10% FBS and cultured in T25 culture flask at 37°C in 5% CO₂-95% air.

Cell traction force microscopy.

Renal VSMCs were grown on flexible substrate impregnated with fluorescent microspheres. Cell traction force exerted on the substrate (by individual VSMC) was determined from measured substrate displacements using the LIBTRC analysis library developed by Micah Dembo of the Boston University, who is also the inventor of the basic theory that underlies traction force microscopy (8, 24, 25). In brief, polyacrylamide gels (5% acrylamide, 0.1% bis-acrylamide, 75-μm thickness, and Young's modulus = 28 kN/m²) were prepared on glass coverslips with fluorescent microspheres (0.2-μm diameter, yellow-green FluoSpheres; Invitrogen) impregnated under the top surface. The Young's modulus of polyacrylamide sheets was determined with an improved method based on the Hertz theory (4, 23, 26) and confirmed with atomic force microscopy. The surface was then coated with collagen type I (0.2 mg/ml; BD Biosciences) after photoactivation of the surface molecular linker, sulfosuccinimidyl-6-(4′-azido-2′-nitrophenylamino) hexanoate (Sulfo-SANPAH; Pierce Chemical) (2), which reacted nonspecifically to the polyacrylamide gel on one end and to the primary amines on the other end, linking the collagen to the polyacrylamide gel. Renal VSMCs were then seeded and cultured overnight. Fluorescent images of the impregnated microspheres right under the cell of interest were captured with a cooled CCD camera (MicroMax:512BFT; Princeton Instruments) mounted on a Zeiss microscope (Axiovert 200M) equipped with an environmental chamber. Images were collected with a ×20 objective lens (Plan-Apochromat, N.A. 0.75) at 37° C. To improve the temporal resolution in measuring cell traction force, fluorescent images and transmitted light images in some experiments were collected simultaneously with a Leica TCS SP5 confocal imaging system at 37° C.

To measure the cell traction force, the image of the cell of interest and the impregnated fluorescent microspheres underneath was first collected, followed by capturing a single image or a series of images during and after the cell was exposed to experimental maneuvers. The cell was then removed from the flexible substrate, and an image of the fluorescent microspheres without the cell adhered was collected. The last image provides the location reference for each fluorescent microsphere when there is no cell traction force. LIBTRC was then applied to the images to tract the displacement of microspheres and to compute the cell traction force. Before the availability of LIBTRC for in-house image processing, images of traction force microscopy were processed and analyzed in Boston University.

Application of electromechanical force on renal VSMCs.

Tosyl-activated paramagnetic microbeads (2.8- or 4.5-μm diameter; Dynal) were coated with fibronectin (Sigma), low-density lipoprotein (LDL; Sigma), or β₁-integrin antibody (Ha2/5; Pharmingen) using the protocol provided by the manufacturer. Only cells in solitude were used for cell traction force study. After the initial cell image without beads was acquired, ligand-coated or noncoated paramagnetic beads were added to the cell chamber and incubated with VSMCs for 45 min to allow the formation of focal adhesions. The prepull image was then collected before turning on the electromagnet. The cell traction force measured from the prepull image was used as the baseline to normalize the cell traction force measured in the same experiment/time series. Controlled force was applied to VSMCs by attracting the paramagnetic beads with a miniaturized needle-shaped electromagnet (tip diameter of 420 μm, 100–150 μm away) mounted on a manipulator as described and characterized previously (1). The force applied to each cell was in the range of 0.5–2 nN. Images were then collected during and after the pulling process. Fibronectin, the predominant native ligand for α₅β₁-integrins in VSMCs (31, 40, 46), was used to examine the integrin-dependent response, and an LDL-coated bead was used as the control for nonintegrin-mediated mechanical signal transduction. Cell surface receptors can transmit mechanical stresses to extracellular matrix at focal adhesions (19).

Intracellular Ca²⁺ concentration measurement in renal VSMCs.

Renal VSMCs grown on collagen-coated Petri dishes with glass bottoms were loaded with fluo-4/AM (10 μM; Invitrogen) in HBSS (Mediatech) for 30 min, followed by 20 min for deesterification of the dye. Fluo-4 was excited at 488 nm, and its emission was collected with a bandpass filter at 522/35 nm. Fluorescent images were collected with either a Bio-Rad MRC-1000 confocal unit coupled to a Zeiss Axiovert 100TV inverted microscope or a Leica TCS SP5 confocal imaging system using water immersion objective lens (×63, N.A. 1.2). Global Ca²⁺ mobilization in VSMCs was monitored at 0.5 Hz in the presence of an activating antibody specific to β₁-integrin (50 μg/ml, Ha2/5, hamster IgM; Pharmingen). Subcellular Ca²⁺ transience was monitored at 500 Hz (2 ms/scan line) in XT line scanning mode.

Immunohistochemistry.

The β₁-integrin on renal VSMCs was first activated with the activating antibody (50 μg/ml, Ha2/5) for 20 min, then fixed with 4% paraformaldehyde in PBS for 20 min, and permeabilized with 0.2% Triton-X in PBS for 30 min. The cells were then incubated with an antibody specific to activated β₁-integrin (10 μg/ml, HUTS-4, mouse IgG; Chemicon) overnight at 4°C, followed by incubation of DyLight 549-conjugated secondary antibodies (3 μg/ml; Jackson ImmunoResearch) for 2 h. An isotype of hamster IgM antibody (50 μg/ml; Santa Cruz) was used as control. Confocal fluorescence images were collected with a water immersion objective lens (×63, N.A. 1.2).

Data analysis.

Normalized cell traction force was reported as means ± SE. Statistical significance (P < 0.05) was assessed by paired or unpaired Student's t-test whenever applicable. Ca²⁺ sparks were detected in confocal images with a custom-made algorithm written in Interactive Data Language software (IDL; Research Systems, Boulder, CO). The program identified Ca²⁺ sparks on the basis of their statistical deviation from the background noise (5, 47). The duration and width of Ca²⁺ sparks were quantified as the full duration half-maximum and full-width half-maximum, respectively (36). The spark frequency of each cell was defined as the number of sparks detected per second in a scan line of 25 μm.

The time series of fluo-4 emission variations in individual renal VSMCs was sampled at 0.5 Hz for spectral analysis. Each time series was normalized with respect to the baseline before β₁-integrin activation and was subjected to linear trend removal. The 256 data points were used to calculate the power spectrum with an algorithm based on Fast Fourier Transform (44). Mean normalized power spectrum was reported as means ± SE.

RESULTS

Nonintegrin-mediated mechanotransduction in renal VSMCs.

To study myogenic response at the single cell level, cell traction force was determined in renal VSMCs grown on a deformable polyacrylamide substrate impregnated with fluorescent microspheres before, during, and after mechanical force was applied by dragging cell-adherent paramagnetic beads under an electromagnetic field generated by a pointed-tip source positioned 100–150 μm away. Under resting conditions, the averaged two-dimensional cell area was 23.7 ± 16 × 10⁻⁶ cm², the total cell traction force was 0.20 ± 0.04 dyn/cell, and normalized cell traction force was 8,260.6 ± 1,749.3 dyn/cm² (1 Pascal = 1 N/m² = 10 dyn/cm²), with a ratio of force along the long axis vs. short axis of 2.51 ± 0.26 (n = 39 cells).

To monitor changes in cell traction forces induced by mechanical force, changes in cell traction force were monitored before bead attachment (no bead), after bead attachment (prepull), 30 s after initiation of the pulling process (pull), and 30 s after the termination of the pulling process (after pull). Attachment of fibronectin beads resulted in a drop of cell traction force, because the formation of new dorsal adhesion sites and redistribution of force (Fig. 1). The decrease in cell traction force was less with LDL beads than with fibronectin beads (19), because LDL beads did not form focal adhesion via integrin. Electromagnetic attraction of noncoated beads caused a small increase in traction force (12 ± 9%, nonsignificant; n = 8 cells), which disappeared when the pulling process was terminated (Fig. 1). Force maps showing the distribution, magnitude, and the direction of the traction force before, during, and after a renal VSMC was pulled with noncoated beads are shown in Fig. 2, A–C. The nonspecific adhesion was apparently ineffective for transducing the applied mechanical force into the VSMCs, presumably due to the lack of strong specific interactions between the cell and the paramagnetic beads.

Fig. 1.

Changes in mean normalized total cell traction force before and after attachment of paramagnetic beads to renal vascular smooth muscle cells (VSMCs), and when renal VSMCs were pulled with fibronectin-coated (FN-coated), LDL-coated, and noncoated paramagnetic beads. Cell traction force was normalized with the traction force measured after paramagnetic beads attached (prepull). The mean total cell traction force is 0.28 ± 0.09 dyn/cell at prepull (24 cells). Cell traction force was restored to prepull level after the pulling process was terminated except in cells pulled by fibronectin-coated paramagnetic beads. Error bars are SE. *Significant difference compared with the prepull control (P < 0.05).

Fig. 2.

Force maps of a renal VSMC pulled by uncoated paramagnetic beads (A, B, and C), LDL-coated paramagnetic beads (D, E, and F), and fibronectin-coated paramagnetic beads (G, H, I). Traction forces were measured before pulling (A, D, and G), 30 s after initiation of pulling (B, E, and H), and 30 s after termination of pulling (C, F, and I). Scale bars = 90 μm. Arrows indicate the direction of pulling force and the position of the electromagnet. Colored scales of cell traction force are in Pascal.

In contrast, electromagnetic pulling of LDL-coated beads increased cell traction force by 61 ± 12% (Fig. 1; n = 9 cells). The cell traction force returned to the prepull level after the pulling process was terminated, and no remanent cell traction force was detected. The force maps of a cell pulled with LDL-coated beads are shown in Fig. 2, D–F. LDL beads were more efficient than noncoated bead in transmitting mechanical force to the focal adhesion because LDL beads interacted with LDL receptors on plasma membrane.

Integrin-mediated mechanotransduction.

To examine the possible role of integrin in the myogenic response, fibronectin-coated paramagnetic beads were used to transduce mechanical force via α₅β₁-integrins. Pulling of fibronectin-coated beads increased the cell traction force by 56 ± 20% (Fig. 1; n = 7 cells), which was similar to the effects of pulling LDL-coated beads. However, the cell traction force did not return to the prepull level after the electromagnet was turned off and moved away. The cell traction force remained elevated at 23 ± 14% (n = 7 cells) of the prepull level. The force maps of a cell pulled with fibronectin-coated beads are shown in Fig. 2, G–I. The elevated cell traction force resembled the changes of vascular resistance in pressure induced myogenic response, in which vascular resistance remains elevated after myogenic constriction.

To further examine whether activation of α₅β₁-integrins with mechanical stimuli can account for the effects on cell traction force elicited by fibronectin-coated beads, renal VSMCs were pulled with paramagnetic beads coated with β₁-integrin antibody. Attachment of β₁-integrin antibody-coated beads resulted in a reduction of cell traction force as observed in fibronectin beads (Fig. 3C), which suggested that beads attachment alone did not activate integrins and generate cell traction force. The changes of cell traction force were monitored continuously during and after the pulling process. Application of mechanical force caused an immediate increase in cell traction force by 40 ± 6% (Fig. 3A; n = 6 cells), when the cells were pulled with the antibody-coated beads, and remanent cell traction force was observed after the pulling process was terminated. Application of mechanical force via beads coated with control isotype antibody only caused a transient increase of cell traction force, which returned to prepull level after the pulling process was terminated (Fig. 3B). The summary of the mean normalized data was shown in Fig. 3C. The nonspecific adhesion between VSMCs and isotype antibody-coated beads was ineffective for transducing applied mechanical force into the VSMCs as observed in noncoated beads. These observations suggest that mechanical activation of β₁-integrin might account for the remanence of cell traction force, consistent with the notion that integrin-mediated mechanotransduction is an integral part of the myogenic response in VSMCs. The rhythmic variations of cell traction force induced by β₁-integrin antibody-coated beads are shown in Fig. 4.

Fig. 3.

Mean time courses of changes in cell traction force induced by pulling paramagnetic beads coated with β₁-integrin antibody (A), paramagnetic beads coated with control isotype antibody (B), and soluble β₁-integrin antibody alone without paramagnetic beads (D). Time courses from individual cells are shown in insets. Horizontal lines indicate duration of pulling. Dotted lines are ± SE. Cell traction force was normalized with the traction force measured after paramagnetic beads attached (prepull). Mean normalized data are shown in histograms (C). Solid bars and opened bars are β₁-integrin antibody and control isotype antibody, respectively. Error bars are SE. *Significant difference compared with the prepull control (P < 0.05, n = 6 cells). The β₁-integrin antibody was added to the bath at t = 0 at (D). Preincubation of Src inhibitor PP2 (5 μM) abolished the increase of cell traction force induced by soluble β₁-integrin antibody ( n = 6 cells).

Fig. 4.

Rhythmic variations of remanent cell traction force. A sequence of consecutive force maps (1–14) sampled at 0.2-Hz display rhythmic variations in cell traction force induced by pulling with β₁-integrin antibody-coated beads. Images shown were collected after the pulling process was terminated. Cell traction force in the colored scale is 1.35E + 04 Pascal. Scale bar = 50 μm.

To examine whether activation of β₁-integrins is causally linked to generation of cell traction force, cell traction force was monitored when VSMCs were incubated with activating β₁-integrin antibody (Ha2/5, 50 μg/ml). Activation of β₁-integrins triggered a sustained increase of cell traction force, which was abolished by pretreatment of the Src inhibitor 4-amino-5-(4-chlorophenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine (PP2; Fig. 3D). It has been shown that mechanical force transduced by β₁-integrin (pulling of magnetic beads coated with activating β₁-integrin antibody) activates Src in smooth muscle cells at real time (33). Taking together, these observations suggest that activation of β₁-integrins with antibody and mechanical force transduced by β₁-integrins share a common signaling pathway to generate cell traction force.

Activation of β₁-integrins and Ca²⁺ mobilization.

It is interesting to note that the cell traction force is oscillatory during and after mechanical stimulation using the β₁-integrin antibody-coated beads. Since increase of intracellular Ca²⁺ concentration is a major determinant of myogenic contraction (29), the kinetic properties of β₁-integrin-mediated Ca²⁺ mobilization was further examined in renal VSMCs. The activating antibody specific to β₁-integrin (Ha2/5, 50 μg/ml) was used to stimulate β₁-integrin and its effect on intracellular Ca²⁺ was monitored with confocal fluorescence microscopy. Activation of β₁-integrin triggered a large initial Ca²⁺ spike followed by periodic Ca²⁺ oscillations (Fig. 5A). Power spectrum analysis from 45 VSMCs showed that most of the Ca²⁺ oscillations occurred at a frequency <0.03 Hz. (Fig. 5B), which is comparable for the rhythmic variations found in cell traction force (Fig. 4). The Ca²⁺ mobilization induced by activating the β₁-integrin antibody was also abolished by SRC inhibitor PP2 (Fig. 5, A and B) as observed in cell traction force study (Fig. 3D).

Fig. 5.

Ca²⁺ mobilization in renal VSMCs induced by soluble β₁-integrin antibody. Normalized time courses of fluo-4 emission in a single cultured renal VSMC (A) and the mean normalized power spectra of oscillations in fluo-4 emission (B). The β₁-integrin antibody was added to the bath at t = 0. Solid lines are mean normalized power spectra (45 cells/3 experiments). Dotted lines are ± SE. Preincubation of Src inhibitor (5 μM PP2) attenuates β₁-integrin antibody induced Ca²⁺ oscillations.

In addition to global Ca²⁺ mobilization, local subcellular Ca²⁺ release (Ca²⁺ sparks) triggered by β₁-integrins was also monitored in freshly isolated renal VSMCs using confocal fluorescence microscopy at 500 Hz in linescan mode. It was shown previously that mechanical force transduced by integrin triggers Ca²⁺ sparks in renal VSMCs (1). Ca²⁺ sparks can initiate or modulate VSMCs contractility via Ca²⁺-activated channels on the plasma membrane. Application of β₁-integrin antibody (Ha2/5) to VSMCs enhanced the occurrence of Ca²⁺ sparks (Fig. 6A) compared with the control using β₂-integrin antibody (mouse, monoclonal IgG, Pharmingen, 50 μg/ml). Since β₂-integrin is not expressed in VSMCs, the Ca²⁺ sparks detected in the presence of β₂-integrin antibody could be considered as spontaneous sparks. Similarly, stimulating the α₅-integrin with antibody (HMα5–1, 25 μg/ml; Pharmingen) could also trigger Ca²⁺ sparks as predicted, because α₅β₁-integrin heterodimers are the predominant fibronectin receptors in VSMCs.

Fig. 6.

Occurrence and characterization of Ca²⁺ sparks induced by soluble integrin antibody. XT line scan image of a single VSMC exhibiting Ca²⁺ sparks (white arrowheads) in the presence of β₁-integrin antibody (A). The whole image is composed of 512 scan lines collected at 500 Hz. Distribution of full width at half the maximum (FWHM, amplitude) and full duration at half maximum (FDHM, duration) of Ca²⁺ sparks in the presence of α₅-integrin antibody (B and C, 83 Ca²⁺ sparks), β₁-integrin antibody (D and E, 71 Ca²⁺ sparks), and β₂-integrin antibody (F and G, 33 Ca²⁺ sparks) were similar. Medians of FWHM are 0.9, 1.2, and 0.8 μm, respectively. Corresponding medians FDHM are 20, 22, and 18 μs.

The mean Ca²⁺ spark frequency in the presence of α₅-, β₁-, and β₂-integrin antibodies were 0.5 ± 0.07 sparks/s (n = 36 cells), 0.6 ± 0.08 sparks/s (n = 35 cells), and 0.04 ± 0.02 sparks/s (n = 30 cells). The frequency of sparks triggered in renal VSMCs by α₅- and β₁-antibodies was significantly higher (P < 0.05) than those triggered by β₂-integrin antibody. The frequency distribution of the spatiotemporal parameters of Ca²⁺ sparks triggered by α₅-, β₁-, and β₂-integrin antibodies is shown in Fig. 6.

Integrins can exist in active or inactive state. Figure 7 is the immunofluorescence of activated β₁-integrin in freshly isolated renal VSMCs with and without activation by Ha2/5 antibody. An isotype of hamster IgM antibody was used as the control for Ha2/5 antibody. Immunofluorescence of β₁-integrin in activated form was enhanced on the plasma membrane after exposure to Ha2/5 antibody but not in the control. The mean immunofluorescence intensity of β₁-integrin in the activated form was increased by 110 ± 8% (3 experiments, 42 cells) after activation compared with the control.

Fig. 7.

Immunofluorescence of activated β₁-integrin detected with antibody HUTS-4 in freshly isolated renal VSMCs after activation by antibody Ha2/5 (A) and control isotype antibody (B). Corresponding transmitted light images are shown in C and D, respectively. Strong immunofluorescence of activated β₁-integrin was localized on the plasma membrane after activation. Scale bars in C and D = 10 μm. Mean immunofluorescence intensity was increased by 110 ± 8% (3 experiments, 42 cells) after activation compared with the control.

DISCUSSION

Myogenic response and tubuloglomerular feedback are the two major mechanisms to autoregulate renal blood flow to prevent the fluctuations of arterial blood pressure from being transmitted into the kidneys. The sensor for tubuloglomerular feedback resides on the macula densa of distal tubule in each nephron, which senses the flow-dependent changes in the luminal NaCl concentration. However, the sensing mechanisms for detecting the changes of transmural pressure in myogenic response are still elusive. Interactions between integrins and extracellular matrix might contribute to this mechanotransduction process. It was previously demonstrated that the magnetic pulling of fibronectin-coated paramagnetic beads in renal VSMCs triggers intracellular Ca²⁺ release from ryanodine receptors as Ca²⁺ sparks (1). Coupling of local Ca²⁺ sparks to adjacent Ca²⁺-activated Cl⁻ channels might lead to membrane depolarization, Ca²⁺ influx, and myogenic response (22). However, direct evidence establishing the link between integrin-mediated mechanotransduction and myogenic response is still missing. In the present study, cell traction force microscopy was used to monitor the contractile reaction of the entire cell when mechanical force was applied via integrin-dependent and integrin-independent adhesion and when β₁-integrin was activated with antibody. Magnetic pulling of fibronectin-coated paramagnetic beads elicited an immediate sizable increase of cell traction force, indicating efficacious transduction of force from paramagnetic beads to the cells. Moreover, the cell traction force remained elevated after the pulling process was terminated. The remanent cell traction force is not due to the residual magnetization of the paramagnetic beads, which are not ferromagnetic. It was not observed when noncoated beads or LDL-coated beads were used, suggesting that the remanent cell traction force is specific to integrin-mediated mechanotransduction. This persistently sustained cell traction force was also found when paramagnetic beads were coated with β₁-integrin antibodies. Addition of fibronectin-coated beads and β₁-integrin antibody-coated beads to cell induced a small decrease in the cell traction force instead of an increase of cell traction (Figs. 1 and 3C), which suggested that the beads attachment alone did not activate integrin and generate cell traction force. By using atomic force microscopy to directly measure and apply nanoscale force to live cultured VSMCs via fibronectin-coated microbeads, Sun et al. (38) demonstrated that upward mechanical force applied to fibronectin-integrin adhesion sites induces an actin-dependent, myogenic response-like counteracting downward pull. Activation of β₁-integrins with antibody triggered a Src-dependent increase of cell traction force. It is known that mechanical force transduced by β₁-integrin activates Src in smooth muscle cells at real time (33). Taken together, these observations suggest that activation of β₁-integrins and mechanical force transduced by β₁-integrin share a common signaling pathway to generate cell traction force.

Integrins exist in both the active and inactive state. Ligand recognition has been shown to induce conformational change in integrins, leading to the exposure of neo-epitopes in ligand-induced binding sites (12, 13). A number of monoclonal antibodies have been shown to activate integrins directly, and a subset of this class of antibodies can recognize ligand-induced binding sites (13, 18). These antibodies can be used to activate specific integrins and to differentiate the active and inactive states of integrins. In this study, the functional antibody Ha2/5 was used to activate β₁-integrins, and β₁-integrin in activated conformation after Ha2/5 incubation was successfully detected with antibody HUTS-4, confirming our experimental conditions for β₁-integrin activation. It is noteworthy that activation of β₁-integrins triggered periodic Ca²⁺ oscillations in renal VSMCs. By sampling cell traction force at 0.2 Hz, the oscillatory remanent cell traction force triggered by β₁-integrin-coated bead was detected. The increase of intracellular Ca²⁺ is a hallmark of myogenic response (30). Hence, the periodic Ca²⁺ spikes and oscillatory remanent cell traction force induced by the activation of β₁-integrins are consistent with this general consensus of myogenic response, as well as the oscillatory nature of myogenic response in renal arterioles (6, 45). Myogenic response in the renal microcirculation of the rat kidney is oscillating at 0.1–0.2 Hz, which is an order of magnitude faster than that of cell traction force oscillation. The kinetic difference of myogenic response in intact renal arterioles and renal VSMCs in culture is mostly likely due to the difference of experimental conditions.

The mechanism for integrin-induced Ca²⁺ mobilization is unclear, but insights might be gained from the local Ca²⁺ signaling at the subcellular level. Magnetic pulling of fibronectin-coated beads triggers Ca²⁺ sparks in freshly isolated renal VSMCs (1). This process is independent of Ca²⁺ influx since removal of extracellular Ca²⁺ did not inhibit Ca²⁺ sparks induced by the pulling force exerted via fibronectin-coated paramagnetic beads (1). Even though it has been shown that Ca²⁺ influx via L-type Ca²⁺ channels contributes to global cytosolic Ca²⁺ in myogenic response, which in turn slowly increases luminal sarcoplasmic reticulum Ca²⁺ load and Ca²⁺ sparks activity (11), this process is not closely coupled to the integrin-dependent Ca²⁺ sparks activation in rat renal VSMCs. Moreover, pulling of fibronectin-coated beads or activation of β₁-integrins with antibodies triggered Ca²⁺ sparks within a few seconds. Hence, it is unlikely that the increase in Ca²⁺ spark activity is the result of a slow increase of sarcoplasmic reticulum Ca²⁺ load. Alternatively, integrin-dependent Ca²⁺ spark activation could be related to cADP-ribose, an endogenous agonist of ryanodine receptors. It has been shown that application of local mechanical force via integrins in smooth muscle activates Src rapidly (<0.3 s) at a remote cytoplasmic site (39), and activation of Src can stimulate ADP-ribose cyclases for cyclic ADP-ribose synthesis (16). This notion is supported by our previous finding in pulmonary VSMCs that the integrin ligand RGD-peptide causes a rapid increase in the level of cyclic ADP-ribose (40). In addition, stretch-induced Ca²⁺ sparks can be mediated by NO via S-nitrosylation of ryanodine receptors (35, 37). Since integrin stimulation for example by RGD peptides induced phosphorylation of focal adhesion kinase, intracellular Ca²⁺ release, and NO production (41), integrins might enhance Ca²⁺ release via FAK/NOS pathways.

In summary, mechanical force applied via the fibronection-α₅β₁-integrin interactions not only increased the cell traction force of renal VSMCs during but also after the removal of the stimulus. The remanent cell traction force was specific to integrin-mediated mechanotransduction, which was not found with LDL-coated beads. Direct activation of β₁-integrins with soluble antibodies triggered periodic Ca²⁺ spikes and increase of cell traction force, which might account for the oscillatory cell traction force induced by mechanical pulling with β₁-integrin antibody-coated beads. In conclusion, mechanical force transduction by α₅β₁-integrins triggers a myogenic response-like behavior in isolated renal VSMCs.

GRANTS

This study was supported by National Institutes of Health Grants 1R03-CA-123621-01A1 (to C.-M. Lo) and HL-071835 (to J. S. K. Sham) a Grant-In-Aid (toK.-P. Yip) from the American Heart Association Greater Southeast.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: L.B., C.-M.L., J.S.S., and K.-P.Y. conception and design of research; L.B., C.-M.L., and K.-P.Y. performed experiments; L.B., C.-M.L., J.S.S., and K.-P.Y. analyzed data; L.B., C.-M.L., J.S.S., and K.-P.Y. interpreted results of experiments; L.B., C.-M.L., and K.-P.Y. prepared figures; L.B. and K.-P.Y. drafted manuscript; L.B., C.-M.L., J.S.S., and K.-P.Y. edited and revised manuscript; L.B., C.-M.L., J.S.S., and K.-P.Y. approved final version of manuscript.