J Physiol Volume 508, Number 1, 187-198, April 1, 1998
The Journal of Physiology (1998), 508.1, pp. 187-198
© Copyright 1998 The Physiological Society
Cytoskeletal targeting of calponin in differentiated, contractile smooth muscle cells of the ferret
Christopher A. Parker * ¹, Katsuhito Takahashi ², Jay X. Tang ³, Terence Tao ¹ ¶ and Kathleen G. Morgan * ¹
* Cardiovascular Division, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston MA 02215, USA, ¹ Boston Biomedical Research Institute, Boston, MA 02114, USA, ² Department of Medicine, Osaka Medical Centre for Cancer and Cardiovascular Diseases, Osaka 537, Japan, ³ Brigham & Women's Hospital, Boston MA 02115, USA and ¶ Department of Biochemistry, Tufts University School of Medicine, Boston, MA 02111, USA
Received 14 August 1997; accepted after revision 1 December 1997.
 |
ABSTRACT |
- Biochemical and quantitative image analysis methods were used to investigate the anatomical basis for the previously described agonist-induced redistribution of calponin.
- At 140 nm resolution, the quantitative distribution of calponin in resting cells was statistically indistinguishable from that of filament bundles containing
-smooth muscle actin and myosin, but was significantly different from that of filaments containing 
-non-muscle actin. Conversely, in stimulated cells, the distribution of calponin was not significantly different from that of -actin filaments in the subplasmalemmal cell cortex but was significantly different from the distribution of -actin- and myosin-containing filamentous bundles.
- The distribution of calponin significantly differed from that of the intermediate filament proteins vimentin and desmin as well as that of the dense body protein -actinin either by ratio analysis of the subcellular distribution or by colocalization analysis.
- The imaging results, although limited to 140 nm spatial resolution, suggested the hypothesis that the agonist-induced redistribution involves the binding of calponin to isoform-specific actin filaments. This hypothesis was tested by quantifying the relative affinity of calponin for purified - and -actin. Light scattering measurements showed that calponin induces bundle formation with -actin more readily than -actin, indicating that calponin may be preferentially sequestered by -actin under appropriate conditions.
- These results are consistent with a model whereby agonist activation decreases calponin's binding to filaments, but the tighter binding to -actin filaments results in a spatial redistribution of calponin to the submembranous cortex.
|
INTRODUCTION |
It is clear that phosphorylation of the 20 kDa myosin light chains (LC20) represents a major pathway by which contractility of differentiated smooth muscle cells is regulated. Evidence, however, is growing for the presence of additional mechanisms regulating smooth muscle contraction, particularly involving thin filament-associated proteins such as caldesmon (CaD) and calponin (CaP). CaP exists in three isoforms: basic (h1), neutral (h2) and acidic (Gimona & Small, 1995; Masuda et al. 1996). Basic CaP is an elongated molecule consisting of a single polypeptide chain with a molecular mass of 32-33 kDa (Stafford, Mabuchi, Takahashi & Tao, 1995). A longer, acidic form of CaP of 36 kDa has been isolated from cultured smooth muscle cells (Applegate, Feng, Green & Taubman, 1994). Basic CaP has been reported to be confined to contractile, differentiated smooth muscle and to be associated, in varying proportions, in various studies, with thin filaments, intermediate filaments, and dense bodies (Walsh, Carmichael & Kargacin, 1993; Parker, Takahashi, Tao & Morgan, 1994; North, Gimona, Cross & Small, 1994a; North, Gimona, Lando & Small, 1994b; Mabuchi, Li, Tao & Wang, 1996).
In vitro studies have demonstrated actin binding and myosin ATPase inhibitory activities of CaP (reviewed in Gimona & Small, 1995) and inhibition of filament sliding in an in vitro motility assay (Pohl, Winder, Allen, Walsh, Sellers & Gerthoffer, 1997). CaP can be phosphorylated in vitro by either protein kinase C (PKC) or Ca2+-calmodulin-dependent kinase II (Nakamura, Mino, Yamamoto, Naka & Tanaka, 1993; Winder, Allen, Fraser, Kang & Kargacin & Walsh, 1993). Phosphorylated CaP is neither capable of binding to actin nor capable of inhibiting actomyosin ATPase. Exogenously added CaP is capable of inhibiting contractile activity of permeabilized smooth muscle (reviewed in Horowitz, Menice, LaPorte & Morgan, 1996b). Peptides from the CaP sequence have been reported to reverse an in situ effect of CaP to suppress vascular tone (Itoh, Suzuki, Watanabe, Mino, Naka & Tanaka, 1995; Horowitz, Clement-Chomienne, Walsh, Tao, Katsuyama & Morgan, 1996a) and extraction of CaP results in contraction of smooth muscle cells (Malmqvist, Trybus, Yagi, Carmichael & Fay, 1997). Thus, current evidence points to the possibility that CaP is a physiologically important regulator of contractility of differentiated smooth muscle.
We have previously reported that in freshly enzymatically isolated ferret portal vein (FPV) smooth muscle cells, agonist-induced contractions are accompanied by a protein kinase C-dependent redistribution of CaP away from the core of the cell (Parker et al. 1994). In the current study, we used biochemical and quantitative image analysis methods to investigate the anatomical basis for this effect. Fluorescence microscopy enhanced by deconvolution methods was used rather than electron microscopy because of the enhanced preservation of native protein distribution and antigenicity of antibodies resulting from the milder fixation and lesser tissue handling required by fluorescence microscopy. Our results are consistent with a model whereby agonist-induced signalling events trigger a decrease in the association of CaP with -actin-containing filaments in the actomyosin domain and a consequent relative increase in the association of CaP with -actin-containing filaments in the cell's surface cortex.
|
METHODS |
Cell isolation
Ferrets were killed with chloroform in a ventilation hood according to procedures approved by the Institutional Animal Care and Use Committees of both the Boston Biomedical Research Institute and the Beth Israel Deaconess Medical Center. The portal vein was quickly removed to a dissection dish filled with oxygenated Krebs solution at 22°C. The tissue was cleaned of connective tissue and opened longitudinally. The endothelium was removed by gently rubbing the vessel interior with a blunt probe. Cells were then isolated using a procedure that is a modification of that previously described and developed specifically to retain pharmacological responsiveness and contractility of the freshly isolated cells (Defeo & Morgan, 1986). In all experiments, cells were tested to confirm contractile ability and responsiveness to the -agonist phenylephrine. The minimal resting length for cells studied was 100 µm. Average cell diameter at the widest point was 5-7 µm. In general, it was confirmed that roughly 80 % of cells shorten dramatically to phenylephrine, and, as previously shown, the small percentage of cells that do not shorten still display normal signal transduction, as measured by intracellular [Ca2+] transients, with shortening apparently being prevented by especially tight adherence of the cell to the coverslip (Defeo & Morgan, 1986).
Briefly, tissue was digested in a medium containing elastase (grade II, Boehringer Mannheim), collagenase (Worthington) and trypsin inhibitor (Sigma, type II-S, soybean). Cells were allowed to settle by gravity and to adhere to the glass coverslip. We have previously found that the responsiveness and viability of mammalian vascular cells correlates with their ability to stick to glass surfaces, and we therefore selected for healthy cells by using only those firmly attached to the coverslip. Cells were used for experiments within 4 h of isolation.
Immunofluorescence
To prevent excessive cell shortening during agonist stimulation, which would obscure fluorescent images, cells were first placed in a Hanks' solution for 5 min (mM: 137 NaCl, 5·4 KCl, 5·6 dextrose, 4·2 NaHCO3, 0·42 Na2HPO4, 0·44 KH2PO4, 1·4 MgCl2 and 1·0 CaCl2) to which 300 mM sucrose was added to increase tonicity. This procedure has previously been shown not to interfere with the signal transduction in smooth muscle cells, as measured by depolarization and agonist-induced increases in intracellular [Ca2+] and PKC translocation (Khalil, Lajoie, Resnick & Morgan, 1992; Khalil & Morgan, 1993; Khalil, Lajoie & Morgan, 1994; Khalil & Morgan, 1996) and is thought to prevent contraction at the cross-bridge level (Cecchi & Bagni, 1994). Also, this procedure has previously been shown not to affect CaP localization in resting cells or mean cell length in the resting state (Parker et al. 1994). Arner & Hellstrand (1980) have shown that hypertonic solutions themselves can contract rat portal vein at body temperature, but under the conditions of the present experiments (22°C), sucrose caused no detectable shortening and less that 10 % of the maximal isometric force normally produced by phenylephrine, but completely prevented any further force generation by phenylephrine.
Cells were either kept in Hanks' solution or stimulated with 10 µM phenylephrine for 6 min and immediately fixed with 2 % paraformaldehyde in phosphate-buffered saline (pH 7·4). The excess fixative was quenched with 0·1 mM glycine in Hanks' solution containing 1 % bovine serum albumin. The cells were then permeabilized with 0·1 % Triton X-100, blocked with 10 % goat serum, incubated with the primary antibody in 2 % goat serum, washed, incubated with the secondary antibody and washed to remove excess label. All cells were mounted on coverslips with Fluorosave (Calbiochem, San Diego, CA, USA) before analysis. In all cases it was confirmed that, when the procedure was repeated in the absence of the primary antibody, there was no detectable background fluorescence.
The rabbit anti-gizzard CaP polyclonal antibody preparation and characteristics have been previously described (Parker et al. 1994). The affinity-purified anti-CaP antibody was used at a 1:1500 dilution for most of the imaging studies unless indicated otherwise. Experiments were also performed with a rabbit anti-bovine CaP polyclonal antibody at 1:1000 and a mouse anti-human CaP monoclonal antibody from Sigma (clone hCP). Rabbit anti-chicken gizzard desmin and -actinin polyclonal antibodies (Sigma) were used at 1:1000 dilutions. Monoclonal anti--smooth muscle actin (clone no. 1A4, Sigma), anti--actin (clone no. AC-15, Sigma), and anti-smooth muscle myosin (clone no. hSM-V, Sigma) were used at 1:1000 dilutions. Monoclonal anti-vimentin (clone no. VIM-13·2, Sigma) was used at a 1:500 dilution.
Digital image analysis
Images were obtained either on a Zeiss IM35 inverted microscope or a Nikon Diaphot 300 equipped with a Nikon ×100 oil immersion objective (NA 1·3). Filters used were 470 ± 20 nm (excitation), 505 nm (dichroic), and 535 ± 20 nm (emission) for fluorescein isothiocyanate (FITC) labelled cells; 560 ± 20 nm (excitation), 595 nm (dichroic), and 630 ± 30 nm (emission) for Texas Red labelled cells. An automated shutter placed in front of the excitation lamp was opened only for data collection to minimize photobleaching. Images were recorded with a liquid cooled charge-coupled device (CCD) camera (Photometrics CH250 or PXL1) via PMIS image processing software (Photometrics) attached to a MS DOS-based microcomputer. A series of two-dimensional images were collected at 0·1 or 0·4 mm intervals between image planes using a computer-controlled stepper motor. The digitized images were then processed and analysed on a SPARC station 5 computer using deconvolution and analysis algorithms written in the Khoros image processing environment and based on Bayesian Maximal Entropy theory, using a priori information as previously described (Parker et al. 1994). Point spread functions for each microscope's optical settings were obtained using 140 nm-diameter polystyrene beads coated with a fluorescent dye at the same electronic and optical settings used when cellular data were acquired.
Colocalization of pairs of proteins within a single cell were performed using secondary antibodies tagged with either FITC or Texas Red. In all cases it was confirmed that there was no detectable cross-talk between labels by exchanging excitation/emission filters for each fluorochrome. The degree of colocalization was determined by a logical 'AND' operation (Rasure, Argiro, Sauer & Williams, 1990) in order to determine whether labelling for each protein coexisted within each pixel of the image. The percentage of colabelled pixels versus the total pixels labelled for each protein was then calculated and reported as percentage colocalization.
A previously described ratio analysis (Parker et al. 1994) of fluorescence intensities was performed in order to determine the relative distribution of various labelled proteins within each cell and to normalize for possible differences in staining between cells. A ratio (R) for a central optical section was calculated (after deconvolution) by determining the mean pixel intensity (total intensity, divided by the number of pixels) of the outer 20 % of the cell (surface cortex) and dividing this value by the mean pixel intensity of the remaining area (cell core). The part of the section containing the nuclear area was avoided when calculating R values. For orientation, we determined the R value for a relatively homogeneous cytoplasmic image resulting from fura-2 loaded FPV cells collected at 350 ± 5·5 nm, which is near the isosbestic point for this dye. This resulted in an average R value of 0·85 (Table 1).
Table 1. Colocalization analysis
| Percentage colocalization |
| Control | Stimulated |
| CaP coincidence with -actin | 67 | 59 |
| -Actin coincidence with CaP67 | 65 | 35 |
| CaP coincidence with -actin | 41 | 63 |
| Myosin coincidence with CaP | - | 20 |
| CaP coincidence with myosin | - | 37 |
5-10 cells were analysed in each case. Data are the percentage of first protein that colocalizes with second protein.
All quantitative values are presented as means ± S.E.M. and comparison of means was done by ANOVA, with P < 0·05 taken as significant.
Proteins
-Skeletal actin was prepared from rabbit skeletal muscle acetone powder according to Spudich & Watt (1971). The non-polymerizing solution contained (mM): 0·2 CaCl2, 0·2 DTT, 0·5 ATP, 0·5 NaN3 and 4 Hepes, at pH 7·5. Actin was polymerized by adding either MgCl2 to 2 mM, or KCl to 150 mM. Pyrene-labelled actin was prepared by the method of Kouyama & Mihashi (1981). Human platelet -actin was purchased from Cytoskeleton, Denver, CO, USA. Recombinant chicken gizzard -CaP was produced as described in Gong, Mabuchi, Takahashi, Nadal-Ginard & Tao (1993). The purified recombinant CaP can be lyophilized and stored as dry protein powder at -20°C for extended periods of time. The lyophilized powder was dissolved in 8 M urea with 10 mM dithiothreitol (DTT) and then dialysed against a solution of 0·1 M KCl, 5 mM DTT and 50 mM Hepes, at pH 7·5. A concentrated stock solution of up to 100 µM was prepared. The protein iconcentration was determined by optical density, assuming a specific absorbance of 0·74 mg ml-1 cm-1 at 280 nm (Stafford et al. 1995) and 32 kDa as the molecular mass for the recombinant CaP.
Light scattering measurements
A Perkin-Elmer LS-5B luminescence spectrometer was used for the 90 deg light scattering measurements. The incident wavelength was set at 365 nm, and a slightly different emission wavelength of 375 nm was chosen; the slit spectral widths were set at 5 nm for both beams. To detect bundle formation by CaP, small aliquots of a concentrated stock solution of 92 µM CaP were added in sequential steps into 600 µl F-actin solution, held in a high UV transparent 5 mm × 10 mm rectangular cuvette, and the light scattering values recorded.
Sedimentation experiments
For the sedimentation assay, actin was polymerized at 0·5 mg ml-1 or higher concentration, by 150 mM KCl. After overnight polymerization, F-actin was diluted with a solution containing (mM): 50 KCl, 0·2 CaCl2, 0·2 DTT, 0·5 ATP and 4 Hepes, pH 7·5. Then CaP was added so that the final mixtures contained 2·5 µM skeletal or platelet actin and 2 µM CaP. The mixtures in 300 µl aliquot were incubated for 30 min, followed by a high speed centrifugation applying 200000 g for 1 h, using a Beckman (Optima TL) table-top ultracentrifuge. CaP molecules which were bound to F-actin cosedimented with the actin filaments, including large oligomers. Free CaP that remained in the supernatant was collected by chloroform precipitation. Total proteins from both supernatant and pellet were subject to SDS-PAGE (12 % polyacrylamide) for direct comparison of amounts of proteins.
|
RESULTS |
A determination of the resolution resulting from optics and deconvolution algorithms used in the present study revealed that two point sources of light were resolvable down to a separation distance of 140 nm in the x-y plane and 100-200 nm along the z-axis. A 3-dimensional image of a 140 nm fluorescent bead was translated 140 nm in the x-y plane and added to itself. The resulting image appeared as one elongated sphere (Fig. 1A). This test image was then subjected to processing via a previously described deconvolution algorithm based on Bayesian Maximum Entropy theory (Parker et al. 1994). The two beads were resolved when translated greater than 140 nm in the x-y plane (Fig. 1B), and thus the resolution of the optics along the x-axis was at least 140 nm. The step size limitation along the z-axis was 100 nm in the system used for this study. Thus when testing the resolution of the optics along this axis the test image was translated in 100 nm incremental steps. When the bead was translated 100 nm along the z-axis and deconvolved, the original two beads were indistinguishable. The two beads were resolved into two separate objects when the distance between them was 200 nm (Fig. 1D), even though they were indistinguishable in the raw image (Fig. 1C). Thus, the resolution in the z-axis was 100-200 nm.
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 1. Determination of resolution of optics
A, image and 3D histogram of 140 nm fluorescent bead translated 140 nm in the x-y plane and added to itself. B, 3D image of bead shown in A after application of deconvolution algorithm. C, image and 3D histogram of 140 nm fluorescent bead translated 200 nm along the z-axis and added to itself. D, 3D image of bead shown in C after application of deconvolution algorithm.
|
In the present study, we addressed the question of the specific proteins and structures with which CaP associates in the smooth muscle cell. Since CaP is an actin-binding protein, we first determined the isoforms of actin in differentiated FPV cells. Western blot analysis of whole cell homogenates of FPV showed, in each case, single, clear bands for both - and -actin isoforms (data not shown). Using the same antibodies, the distribution of the - and -isoforms of actin in single, freshly isolated cells, both in the control and in the stimulated state was determined. Figure 2A illustrates the distribution of -actin in a resting cell. Continuous bundles of -actin, running parallel to the long axis of the cell, but excluding the nuclear space (right-hand side of Fig. 2A) were observed.
We previously published (Parker et al. 1994) a simple ratio analysis algorithm to quantify the distributions of the proteins between the cell surface cortex and the central core of the cell (see Methods). The distribution of -actin did not appear to qualitatively change when cells were stimulated with 10 µM phenylephrine for 6 min (Fig. 2B). The average R value (Fig. 3) of -actin in resting cells was 0·80 ± 0·04. The R value of -actin in 10 µM phenylephrine (6 min) stimulated cells was 0·87 ± 0·16. These values were not significantly different (P > 0·05), confirming that upon stimulation, -actin does not undergo a significant detectable redistribution in these differentiated smooth muscle cells.
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 2. -Actin localization
A, control FPV cell showing immunolocalization of -actin. B, stimulated FPV cell (10 mM phenylephrine, 6 min) showing immunolocalization of -actin. Two different cells are shown. Scale bar, 5 µm.
|
In contrast, the distribution of CaP changes significantly upon -agonist stimulation. As previously reported (Parker et al. 1994), and confirmed here (see green images in Fig. 5) the R value for CaP in -agonist-stimulated cells increases significantly (P < 0·001) to 1·27 from a resting value of 0·75 (Fig. 3). In the present study, similar results were observed with three different anti-CaP antibodies: an anti-gizzard CaP polyclonal, an anti-bovine CaP polyclonal and an anti-human CaP monoclonal.
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 3. R values resulting from digital image analysis of labelled proteins
For comparison, a homogeneous fura-2 image gives a ratio of 0·85 by this analysis. All bars indicate R values from control cells unless otherwise indicated. At least five cells were analysed for each population.
|
The distribution of the -isoform of actin was found to be qualitatively different from that of -actin. -Actin labelling resulted in a discrete pattern which preferentially occurred in the surface cortex of the cell with a network of small plaques extending throughout the cell (Figs 4A and B and 5). No qualitative redistribution of -actin between resting (Fig. 4A) and stimulated (10 µM phenylephrine) (Fig. 4B) cells was detected. The R value for resting cells labelled for -actin was 1·12 (Fig. 3). This did not differ significantly from the R value found for -actin in stimulated cells, 1·12 ± 0·08 (P > 0·05), but did differ significantly from that found for -actin in resting cells (P < 0·05).
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 4. -Actin localization
A, control FPV cell showing immunolocalization of -actin. B, stimulated FPV cell (10 mM phenylephrine, 6 min) showing immunolocalization of -actin. Two different cells are shown. Scale bar, 5 µm.
|
The R value of -actin was not significantly different from the R value for CaP in resting cells (P > 0·05; Fig. 3) but was statistically significantly different from that for CaP in the stimulated cell (P < 0·001; Fig. 3). Conversely, the R value for -actin was significantly different from that for CaP in the resting cell (P < 0·01; Fig. 3) but not statistically significantly different from that in the stimulated cell (P > 0·05; Fig. 3). Thus, since the distribution of various proteins relative to the whole cell, as indicated by the R values, suggested, but did not prove, an agonist-induced redistribution of CaP from -actin- to -actin-containing filaments, we attempted to test this concept further by quantifying the extent to which the proteins coexisted within a single colabelled FPV cell at the 140 nm resolution of our system. As shown in Fig. 5A, colocalization of CaP and -actin in the same control cell, indicates a high degree of overlap. Quantification of overlapping pixels by a logical 'AND' calculation indicated that on average, of the pixels containing CaP, 67 % also contained -actin, whereas 65 % of the pixels containing -actin also contained CaP (Table 1). After taking into account pixel misregistration, this degree of overlap corresponds to nearly complete overlap of these proteins (Moore et al. 1993). The addition of the -agonist phenylephrine (100 µM, 6 min) caused a dramatic decrease in the degree of colocalization of CaP and -actin (Fig. 5A). In the stimulated cell, a large portion of the -actin in the core of the cell was not associated with CaP. The percentage of the -actin that coincided with CaP decreased to 35 % (Table 1). The percentage of CaP coinciding with -actin remained high (59 %; Table 1) because the surface cortex contained some of the total -actin found within the cell. We cannot resolve if the CaP in the cortex is associated with - or -actin, which intermingle in this domain.
Cells were also colabelled for -actin and CaP in both the resting and stimulated states (Fig. 5B). In the resting state, there is some overlap between CaP and -actin both in the submembranous cortex and in the plaque-like network throughout the cell (41 % of the CaP overlaps with -actin in resting cells; Table 1). Upon stimulation, however, the percentage of CaP that coincides with -actin increased to 63 %, which, considering pixel misregistration artifacts, represents nearly complete colocalization (Fig. 5B and Table 1).
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 5. Colocalization of calponin with -actin or -actin
A, double labelled FPV cells showing immunolocalization of -actin and calponin in the control and stimulated state (10 mM phenylephrine, 6 min). Top row, control cells; bottom row, stimulated cells. Red indicates -actin, green indicates calponin and yellow indicates pixels containing signal for both -actin and calponin. Note that -actin is excluded from the intranuclear space (left-hand side of stimulated cell). B, FPV cells double labelled for -actin and calponin in both the control and stimulated state (10 mM phenylephrine, 6 min). Top row, control cells; bottom row, stimulated cells. Red indicates -actin, green indicates calponin and yellow indicates pixels containing both -actin and calponin. Scale bar, 5 µm.
|
For orientation, myosin and -actinin distribution in this cell preparation were also determined. When entire sets of optical sections from individual cells stained for myosin (Fig. 6A-E) are examined, a pattern emerges of long bundles spiralling through the cell, which is consistent with the reported spiral contraction observed in isolated individual cells (Warshaw, McBride & Work, 1987). In sections from the centre of the cell, myosin ran parallel to the long axis of the cell (Fig. 6C). The mean R value for myosin from a centre optical section was 0·74 ± 0·02. This value was not significantly different from that for either the CaP or -actin in resting cells (P > 0·05) but was significantly less than that for -actin and CaP in stimulated cells (P < 0·01; Fig. 2), a finding consistent with the concept that myosin colocalizes near -actin but not -actin, and also that CaP may exist on myosin-associated actin filaments in the resting state but not in the stimulated state. Indeed, colocalization analysis of CaP and myosin indicates that in stimulated cells, only 20 % of the myosin is colocalized with CaP (Table 1).
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 6. Optical sections through a FPV cell showing immunolocalization of myosin
The sections progress at 0·8 mm steps from A (bottom of the cell) to E (top of the cell). Scale bar, 5 µm.
|
Figure 7 illustrates the typical discrete pattern of -actinin labelling exhibited in FPV cells. It has been suggested that the dense bodies, which contain high concentrations of -actinin, may function in a manner analogous to Z-lines and the interval between dense bodies may represent the smooth muscle equivalent of the sarcomere (Kargacin, Cooke, Abramson & Fay, 1989). The R value for -actinin was 0·82 ± 0·11, a number not significantly different from that for the R for a homogeneous fura-2 image, indicating that -actinin is distributed throughout the cell, and could represent docking sites for both - and -actin. A comparison of Figs 5 and 7 indicates that the distribution of the majority of CaP in resting cells is qualitatively different from the dense body distribution.
The results above, taken together, suggest a model where CaP may preferentially redistribute to areas that contain different isoforms of actin, depending on the activation state of the cell. However, the possibility also remains that CaP may undergo a regulated association with intermediate filaments, since North et al. (1994a) have reported a partial codistribution of CaP with desmin in chicken gizzard cells, and Mabuchi et al. (1996) reported an association with cytoskeletal elements, including desmin, also in chicken gizzard. Western blot analysis of these mammalian vascular cells (data not shown) indicated the presence of two intermediate filament proteins, desmin and vimentin. The distributions of both these proteins in FPV cells was determined, and are illustrated in Figs 8B and 9, respectively. The mean R value (Fig. 3) for desmin was 0·60 ± 0·07 in five resting cells and the R value for five stimulated cells was 0·59 ± 0·03 (P > 0·05), and thus desmin (Fig. 8B) did not appear to undergo any obvious changes in intracellular distribution upon -agonist stimulation. The R value for desmin was not statistically significantly different from that for CaP in the resting cell. However, inspection of individual resting cells colabelled for CaP (Fig. 8A) and desmin (Fig. 8B) consistently showed qualitatively different patterns of staining with the desmin arranged in what appeared to be much shorter bundles than was seen with either CaP (Fig. 8A) or -actin (Fig. 2). Vimentin distribution (Fig. 9) did not resemble that of CaP in resting or stimulated cells. Vimentin was found on average to give an R value (Fig. 2) of 0·97, which was not significantly different from CaP in resting cells (P > 0·05) but clearly lacked the appearance of long continuous bundles of filaments characteristic of CaP resting cells. The R value for CaP in stimulated cells was significantly larger than that for vimentin (P < 0·05; Fig. 3).
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 8. Double labelled unstimulated FPV cell showing immunolocalization of calponin and desmin
A, calponin immunolocalization (Sigma anti-human monoclonal antibody). B, desmin immunolocalization. Scale bar, 5 µm.
|
Thus, the results above, taken together, suggest a model where CaP has differential affinity for actin isoforms. In order to test this idea further, experiments were performed to compare, by cosedimentation and light scattering analysis, the binding of CaP to purified -actin from skeletal muscle with -non-muscle actin from human platelets. Lu, Freedman & Chalovich (1995) have previously shown that there is no difference in CaP binding between -skeletal muscle actin and 
-smooth muscle actin from chicken gizzard. We also found no detectable difference in cosedimentation of CaP with - versus -actin from human platelets (Fig. 10). The cosedimentation experiment was performed at both low (with EDTA) and high (200 µM) free Ca2+ in order to test the additional hypothesis that a difference might occur with the rise in the level of free calcium, but the comparison in Fig. 10 shows no difference independent of the calcium concentration. However, the light scattering experiments clearly showed that the platelet actin filaments are more prone to aggregation by CaP (Fig. 11). The data in Fig. 11 were obtained with F-actin in 50 mM KCl and 2 mM MgCl2 in addition to 200 µM CaCl2. A similar difference was detected without MgCl2, and with addition of 1 mM EGTA to chelate free Ca2+ (data not shown). Therefore, the difference between the two actin isoforms as they were bundled by CaP is independent of the concentrations of free Ca2+ or Mg2+ ions.
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 10. Cosedimentation of CaP with skeletal and platelet actin, with and without 200 µM Ca2+
For the absence of Ca2+, Na2EDTA was added to 1 mM in order to chelate 200 µM Ca2+ that was present in the actin buffer solution. No notable difference was detected between the two isoactins in both high and low free calcium concentrations. See Methods for preparation details.
|
|
 |
View larger version
[in this window]
[in a new window] |
|
|
Figure 11. Comparison in bundle formation of skeletal and platelet F-actin induced by CaP
An increase in light scattering indicates the bundle formation of F-actin-CaP complex. a.u., arbitrary units. See Methods for more details.
|
|
DISCUSSION |
In the present study we have investigated the possible anatomical basis for the -agonist-induced redistribution of CaP in differentiated FPV cells, by localizing CaP relative to cytoskeletal and contractile proteins. The results presented here show that the CaP in the central filamentous bundles of resting cells codistributes with the -smooth muscle isoform of actin as well as myosin. Calponin has been shown to bind actin in vitro (reviewed in Horowitz et al. 1996) and smooth muscle is known to contain both muscle ( and ) and non-muscle ( and ) actin isoforms (Herman, 1993). North, Gimona, Lando & Small (1994b) demonstrated that in chicken gizzard smooth muscle, -non-muscle actin localized in dense bodies and along longitudinal channels that link consecutive dense bodies. This may correspond to the relatively small fraction of -actin staining seen in our study in patches in the core of the cell (Figs 4 and 5). Also, this group reported that in chicken gizzard smooth muscle, -smooth muscle actin exists in filaments devoid of -actin labelling. In contrast, Drew, Moos & Murphy (1991) reported that individual isolated thin filaments from adult swine stomach exhibited a random and mixed distribution of actin isoforms. Our results also differ from those of Drew et al. (1991) in that we found that - and -actin isoforms in the FPV cell display qualitatively different subcellular localizations. The reason for these differences is not clear but the possibility exists that the isolated filaments studied in vitro by Drew et al. (1991) which contained both isoforms represents a subset of total actin filaments. Our results cannot completely rule out the existence of a small subset of filaments where both actin isoforms coexist.
In general, however, like North et al. (1994b), we found that these two isoforms do exist in distinct and separate structures. However, the identification in the differentiated, contractile smooth muscle cell of a submembranous, surface cortex domain, occupied by -non-muscle actin, as seen in the present study, was not reported by North et al. (1994a,b). This network of -actin may be involved in the regulation of plasmalemmal membrane channels and pumps, in signalling complexes, or in subplasmalemmal interactions with integrin complexes, such as has been suggested for actin filaments near the surface membrane in differentiated tracheal smooth muscle (Wang, Pavalko & Gunst, 1996). In contrast, the central core of filaments containing the -actin isoform and myosin represent a contractile domain. The association-dissociation of CaP with this domain in the resting and contracted muscles, respectively, suggests a role for CaP in the regulation of smooth muscle contractility.
In regard to the mechanism of the agonist-induced redistribution of CaP, three hypotheses emerge. First, since these results are based on immunolocalization, CaP in the core of the cell after agonist stimulation may simply no longer be recognized by the antibody used in the study. CaP or associated proteins may undergo a conformational change which sterically hinders the binding of the antibody to CaP, thus resulting in the apparent disappearance of CaP from the core of the cell. This hypothesis seems relatively unlikely since, in the present study, the same redistribution was seen with three different antibodies. A related possibility is that for unknown reasons, only a subset of the isoforms of CaP is recognized in the stimulated state. This possibility, however, is ruled out by the recent report (Masuda et al. 1996) showing that one of the antibodies used in the present study (Sigma monoclonal hCP) is specific for basic (h1) CaP.
A second hypothesis is that signalling pathways specifically regulate the spatial localization of CaP in the 'contractile domain', but not in the subplasmalemmal 'cell cortex'. One possibility is that caldesmon, an actin binding protein that is present in the 'contractile domain' (Furst, Cross, DeMey & Small, 1986) plays a role. CaD has been shown to be phosphorylated upon activation (Adam, Haeberle & Hathaway, 1989; Adam, Gapinski & Hathaway, 1992) in smooth muscle. On cell activation, a conformational change in CaD may inhibit CaP's binding to those actin filaments and thus trigger the redistribution of CaP. We have not ruled out this possibility in the present study, but it is unclear if this mechanism alone could cause a targeting of CaP to the submembranous cortex.
A third hypothesis is that CaP differs in its tendency to associate with - and -actin isoforms and that an agonist-induced decrease in the affinity of CaP for actin results in the removal of CaP from one population of filaments but not the other. Our results are consistent with this hypothesis in that the light scattering data indicate either a higher affinity of CaP for -non-muscle actin, or a subtle difference in the mode of CaP binding to the two actin isoforms. The latter possibility is consistent with a recently proposed electrostatic model for the mechanism of CaP-induced actin bundle formation (Tang & Janmey, 1996; Tang, Szymanski, Janmey & Tao, 1997). If assuming no difference in the binding constant of CaP to - or -actin isoform, the model predicts that the -isoactin with less net negative charge is more likely to bundle than the -actin when the same amount of CaP is bound to the filaments. This bundling effect was detectable with light scattering measurements, but not by cosedimentation assay, which may explain the failure of previous coimmunoprecipitation analysis to observe a difference in CaP's affinity for actin isoforms. It is plausible that, due to the preferential bundling, -isoactin may better sequester the CaP molecules that diffuse away from the contractile region following the cell activation.
The signal for the postulated agonist-induced decrease in actin affinity is still unknown, but the CaP redistribution has been shown to be dependent on protein kinase C (PKC) activity (Parker et al. 1994), suggesting a phosphorylation event. CaP can be phosphorylated by PKC in vitro (Winder et al. 1993). Also, several groups have reported that CaP can be phosphorylated in vivo (Winder et al. 1993; Mino, Yuasa, Naka & Tanaka, 1995; Pohl et al. 1997). Winder et al. (1993) also showed that upon phosphorylation in vitro, CaP's binding affinity to actin is diminished significantly. Phosphorylation may trigger a loss of CaP binding affinity to actin. If CaP has a higher affinity for -actin relative to that for -actin, this may lead to a relatively higher codistribution with -actin. Whether CaP is phosphorylated in vivo, however, remains an open question (reviewed in Gimona & Small, 1995). The in vitro experiments in the present study were performed with unphosphorylated CaP, however, and the possibility remains that CaP itself is not directly phosphorylated but that another signalling molecule is phosphorylated and transmits the PKC-dependent signal to CaP.
Our results also point to the existence of two separate and distinct functional zones in these differentiated cells. The myosin-containing area of FPV cells can be viewed as the contractile, myofilament zone and the myosin free area can be thought of as the non-contractile submembranous cortex. -Actin appears to localize outside of the myosin-containing contractile filament bundles of FPV cells and CaP's apparent redistribution may cause a disinhibition of myosin and contribute to the agonist-induced contraction.
|
REFERENCES |
Adam, L. P., Gapinski, C. J. & Hathaway, D. R. (1992). Phosphorylation sequences in h-caldesmon from phorbol ester-stimulated canine aortas. FEBS Letters 302, 223-226.
[Medline]
Adam, L. P., Haeberle, J. R. & Hathaway, D. R. (1989). Phosphorylation of caldesmon in arterial smooth muscle. Journal of Biological Chemistry 264, 7698-7703.
[Abstract]
Applegate, D., Feng, W., Green, R. S. & Taubman, M. B. (1994). Cloning and expression of a novel acidic calponin isoform from rat aortic vascular smooth muscle. Journal of Biological Chemistry 269, 10683-10690.
[Abstract]
Arner, A. & Hellstrand, P. (1980). Contraction of the rat portal vein in hypertonic and isotonic medium: rates of metabolism. Acta Physiologica Scandinavica 110, 69-75.
[Medline]
Cecchi, G. & Bagni, M. A. (1994). Myofilament lattice spacing affects tension in striated muscle. News in Physiological Sciences 9, 3-7.
Defeo, T. T. & Morgan, K. G. (1986). A comparison of two different indicators: Quin 2 and aequorin in isolated single cells and intact strips of ferret portal vein. Pflügers Archiv 406, 427-429.
Drew, J. S., Moos, C. & Murphy R. A. (1991). Localization of isoactins in isolated smooth muscle thin filaments by double gold immunolabeling. American Journal of Physiology 260, C1332-1340.
[Medline]
Furst, D. O., Cross, R. A., DeMey, J., Small, J. V. (1986). Caldesmon is an elongated, flexible molecule localized in the actomyosin domains of smooth muscle. EMBO Journal 5, 251-257.
[Abstract]
Gimona, M. & Small, J. V. (1995). Calponin. In Biochemistry of Smooth Muscle Contraction, ed. Barany, M, pp. 91-103. Academic Press, San Diego.
Gong, B.-J., Mabuchi, K., Takahashi, K., Nadal-Ginard, B. & Tao, T. (1993). Characterization of wild type and mutant chicken gizzard calponin expressed in E. coli. Journal of Biochemistry 114, 453-456.
Herman, I. M. (1993). Actin isoforms. Current Opinion in Cell Biology 5, 48-55.
[Medline]
Horowitz, A., Clement-Chomienne, O., Walsh, M. P., Tao, T. Katsuyama, H. & Morgan, K. G. (1996a). Effects of calponin on force generation by single smooth muscle cells. American Journal of Physiology 270, H1858-1863.
[Medline]
Horowitz, A., Menice, C. B., LaPorte, R., Morgan, K. G. (1996b). Mechanisms of smooth muscle contraction. Physiological Reviews 76, 967-1003.
[Medline]
Itoh, T., Suzuki, A., Watanabe, Y., Mino, T., Naka, M. & Tanaka, T. (1995). A calponin peptide enhances Ca2+ sensitivity of smooth muscle contraction without affecting myosin light chain phosphorylation. Journal of Biological Chemistry 270, 20400-20403.
[Abstract/Full Text]
Kargacin, G. J., Cooke, P. H., Abramson, S. B. & Fay, F. S. (1989). Periodic organization of the contractile apparatus in smooth muscle revealed by the motion of dense bodies in single cells. Journal of Cell Biology 108, 1465-1475.
[Abstract]
Khalil, R. A., Lajoie, C. & Morgan, K. G. (1994). In situ determination of the [Ca2+]i threshold for translocation of the alpha protein kinase C isoform. American Journal of Physiology 266, C1544-1551.
[Medline]
Khalil, R. A., Lajoie, C., Resnick, M. S. & Morgan, K. G. (1992). Ca2+-independent isoforms of protein kinase C differentially translocate in smooth muscle. American Journal of Physiology 263, C714-719.
[Medline]
Khalil, R. A. & Morgan, K. G. (1993). Protein kinase C-mediated redistribution of mitogen-activated protein kinase during smooth muscle cell activation. American Journal of Physiology 265, C406-411.
[Medline]
Khalil, R. A. & Morgan, K. G. (1996). Enzyme translocations during smooth muscle activation. In Biochemistry of Smooth Muscle Contraction, ed. Barany, M., pp. 307-318. Academic Press, San Diego.
Kouyama, T. & Mihashi, K. (1981). Flourimetry study of N-(1-pyreneyl)-iodoacetamide-labelled F-actin. European Journal of Biochemistry 114, 33-38.
[Abstract]
Lu, F. W. M., Freedman, M. V. & Chalovich, J. M. (1995). Characterization of calponin binding to actin. Biochemistry 34, 11864-11871.
Mabuchi, K., Li, Y.-L., Tao, T. & Wang, L. A. (1996). Immunocytochemical localization of caldesmon and calponin in chicken gizzard smooth muscle. Journal of Muscle Research and Cell Motility 17, 243-260.
[Medline]
Malmqvist, U., Trybus, K. M., Yagi, S., Carmichael, J. & Fay, F. S. (1997). Slow cycling of unphosphorylated myosin is inhibited by calponin, thus keeping smooth muscle relaxed. Proceedings of the National Academy of Sciences of the USA 94, 7655-7660.
[Abstract/Full Text]
Masuda, H., Tanaka, K., Takagi, M., Ohgami, K., Sakamaki, T., Shibata, N. & Takahashi, K. (1996). Molecular cloning and characterization of human non-smooth muscle calponin. Journal of Biochemistry 120, 415-424.
[Medline]
Mino, T., Yuasa, U., Naka, M. & Tanaka, T. (1995). Phosphorylation of calponin mediated by protein kinase C in association with contraction in porcine coronary artery. Biochemical and Biophysical Research Communications 208, 397-404.
[Medline]
Moore, E. D. W., Etter, E. F., Phillipson, K. D., Carrington, W. A., Fogarty, K. E., Lifshitz, L. M. & Fay, F. S. (1993). Coupling of the Na+/Ca2+ exchanger, Na+/K+ pump and sarcoplasmic reticulum in smooth muscle. Nature 365, 657-660.
[Medline]
Nakamura, F., Mino, T., Yamamoto, J., Naka, M. & Tanaka, T. (1993). Identification of the regulatory site in smooth muscle calponin that is phosphorylated by protein kinase C. Journal of Biological Chemistry 268, 6194-6201.
[Abstract]
North, A. J., Gimona, M., Cross, R. A. & Small, J. V. (1994a). Calponin is localized in both the contractile apparatus and the cytoskeleton of smooth muscle. Journal of Cell Science 107, 437-444. &uid=AMBIGUOUS&dopt=r">
[Abstract]
North, A. J., Gimona, M., Lando, Z. & Small, J. V. (1994b). Actin isoform compartments in chicken gizzard smooth muscle cells. Journal of Cell Science 107, 445-455. &uid=AMBIGUOUS&dopt=r">
[Abstract]
Parker, C. A., Takahashi, K., Tao, T. & Morgan, K. G. (1994). Agonist-induced redistribution of calponin in contractile vascular smooth muscle cells. American Journal of Physiology , 267, C1262-1270.
Pohl, J., Winder, S. J., Allen, B. G., Walsh, M. P., Sellers, J. R. & Gerthoffer, W. T. (1997). Phosphorylation of calponin in airway smooth muscle. American Journal of Physiology 272, L115-123.
Rasure, J. D., Argiro, D., Sauer, T. & Williams, C. (1990). A visual language and software development environment for image processing. International Journal of Imaging Systems and Technology 2, 183-199.
Spudich, J. A. & Watts, S. (1971). The regulation of rabbit skeletal muscle contraction. I. Biochemical studies of the interaction of tropomyosin-troponin complex with actin and the proteolytic fragments of myosin. Journal of Biological Chemistry 246, 4866-4871.
[Medline]
Stafford, W. F., Mabuchi, K., Takahashi, K. & Tao, T. (1995). Physical characterization of calponin. Journal of Biological Chemistry 270, 10576-10579.
[Abstract/Full Text]
Szymanski, P. T., Grabarek, Z. & Tao, T. (1997). Correlation between calponin and myosin subfragment 1 binding to F-actin and ATPase inhibition. Biochemical Journal 321, 519-523.
[Medline]
Tang, J. X. & Janmey, P. A. (1996). Polyelectrolyte nature of F-actin and mechanism of actin bundle formation. Journal of Biological Chemistry 271, 8556-8563.
[Abstract/Full Text]
Tang, J. X., Szymanski, P., Janmey, P. A. & Tao, T. (1997). Electrostatic effects of smooth muscle calponin on actin assembly. European Journal of Biochemistry 247, 432-440.
[Abstract]
Walsh, M. P., Carmichael, J. D. & Kargacin, G. J. (1993). Characterization and confocal imaging of calponin in gastrointestinal smooth muscle. American Journal of Physiology 265, C1371-1378.
[Medline]
Wang, Z. L., Pavalko, F. M. & Gunst, S. J. (1996). Tyrosine phosphorylation of the dense plaque protein paxillin is regulated during smooth muscle contraction. American Journal of Physiology 40, C1594-1602.
Warshaw, D. M., McBride, W. J. & Work, S. S. (1987). Corkscrew-like shortening in single smooth muscle cells. Science 236, 1457-1459.
[Medline]
Winder, S. J., Allen, B. G., Fraser, E. D., Kang, H.-M., Kargacin, G. J. & Walsh, M. P. (1993). Calponin phosphorylation in in vitro and in intact muscle. Biochemical Journal 296, 827-836.
[Medline]
Acknowledgements
This work was supported by National Institutes of Health grants HL31704 and HL42293 to K. G. M. The authors wish to acknowledge the help of Mr Justin Hulvershorn in the preparation of the illustrations.
Corresponding author
K. G. Morgan: Boston Biomedical Research Institute, 20 Staniford Street, Boston, MA 02114, USA.
Email: morgan{at}bbri.harvard.edu
This article has been cited by other articles:
|
|
|
|
 
W. Zhang and S. J. Gunst
Interactions of Airway Smooth Muscle Cells with Their Tissue Matrix: Implications for Contraction
Proceedings of the ATS,
January 1, 2008;
5(1):
32 - 39.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. Gallant, J. Y. You, Y. Sasaki, Z. Grabarek, and K. G. Morgan
MARCKS is a major PKC-dependent regulator of calmodulin targeting in smooth muscle
J. Cell Sci.,
August 15, 2005;
118(16):
3595 - 3605.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. S. Gangopadhyay, N. Takizawa, C. Gallant, A. L. Barber, H.-D. Je, T. C. Smith, E. J. Luna, and K. G. Morgan
Smooth muscle archvillin: a novel regulator of signaling and contractility in vascular smooth muscle
J. Cell Sci.,
October 1, 2004;
117(21):
5043 - 5057.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. B. Patil, M. D. Pawar, and K. N. Bitar
Direct association and translocation of PKC-{alpha} with calponin
Am J Physiol Gastrointest Liver Physiol,
June 1, 2004;
286(6):
G954 - G963.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. Samaj, F. Baluska, and H. Hirt
From signal to cell polarity: mitogen-activated protein kinases as sensors and effectors of cytoskeleton dynamicity
J. Exp. Bot.,
January 2, 2004;
55(395):
189 - 198.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. Wrayzx, K. Jones, S. Kupittayanant, Y. Li, A. Matthew, E. Monir-Bishty, K. Noble, S. J. Pierce, S. Quenby, and A. V. Shmygol
Calcium Signaling and Uterine Contractility
Reproductive Sciences,
July 1, 2003;
10(5):
252 - 264.
[Abstract]
[PDF]
|
|
|
|
|
|
|
M. M. Hossain, D.-Y. Hwang, Q.-Q. Huang, Y. Sasaki, and J.-P. Jin
Developmentally regulated expression of calponin isoforms and the effect of h2-calponin on cell proliferation
Am J Physiol Cell Physiol,
January 1, 2003;
284(1):
C156 - C167.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H.-M. Shin, H.-D. Je, C. Gallant, T. C. Tao, D. J. Hartshorne, M. Ito, and K. G. Morgan
Differential Association and Localization of Myosin Phosphatase Subunits During Agonist-Induced Signal Transduction in Smooth Muscle
Circ. Res.,
March 22, 2002;
90(5):
546 - 553.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. E. Cain, D. M. Tanner, and R. A. Khalil
Endothelin-1-Induced Enhancement of Coronary Smooth Muscle Contraction via MAPK-Dependent and MAPK-Independent [Ca2+]i Sensitization Pathways
Hypertension,
February 1, 2002;
39(2):
543 - 549.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
K. G. Morgan and S. S. Gangopadhyay
Signal Transduction in Smooth Muscle: Invited Review: Cross-bridge regulation by thin filament-associated proteins
J Appl Physiol,
August 1, 2001;
91(2):
953 - 962.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y.-H. Lee, C. Gallant, H. Guo, Y. Li, C.-L. A. Wang, and K. G. Morgan
Regulation of Vascular Smooth Muscle Tone by N-terminal Region of Caldesmon. POSSIBLE ROLE OF TETHERING ACTIN TO MYOSIN
J. Biol. Chem.,
February 4, 2000;
275(5):
3213 - 3220.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. Leinweber, A. M. Parissenti, C. Gallant, S. S. Gangopadhyay, A. Kirwan-Rhude, P. C. Leavis, and K. G. Morgan
Regulation of Protein Kinase C by the Cytoskeletal Protein Calponin
J. Biol. Chem.,
December 15, 2000;
275(51):
40329 - 40336.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H.-M. Shin, H.-D. Je, C. Gallant, T. C. Tao, D. J. Hartshorne, M. Ito, and K. G. Morgan
Differential Association and Localization of Myosin Phosphatase Subunits During Agonist-Induced Signal Transduction in Smooth Muscle
Circ. Res.,
March 22, 2002;
90(5):
546 - 553.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Introduction
