Smooth muscle contraction involves phosphorylation of the regulatory myosin light chain. However, this thick-filament system of regulation cannot account for all aspects of a smooth muscle contraction. An alternate site of contractile regulation may be in the thin-filament-associated proteins, in particular caldesmon. Caldesmon has been proposed to be an inhibitory protein that acts either as a brake to stop any increase in resting or basal tone, or as a modulatory protein during contraction. The goal of this study was to use short interfering RNA technology to decrease the levels of the smooth muscle-specific isoform of caldesmon in intact vascular smooth muscle tissue to determine more carefully what role(s) caldesmon has in smooth muscle regulation. Intact strips of vascular tissue depleted of caldesmon produced significant levels of shortening velocity, indicative of cross-bridge cycling, in the unstimulated tissue and exhibited lower levels of contractile force to histamine. Our results also suggest that caldesmon does not play a role in the cooperative activation of unphosphorylated cross bridges by phosphorylated cross bridges. The velocity of shortening of the constitutively active tissue and the high basal values of myosin light chain phosphorylation suggest that h-caldesmon in vivo acts as a brake against contractions due to basally phosphorylated myosin. It is also possible that phosphorylation of h-caldesmon alone in the resting state may be a mechanism to produce increases in force without stimulation and increases in calcium. Disinhibition of h-caldesmon by phosphorylation would then allow force to be developed by activated myosin in the resting state.
- myosin light chain
- shortening velocity
- carotid artery
- cross-bridge cycling
the primary mechanism by which smooth muscle contracts involves calcium/calmodulin-dependent activation of myosin light chain (MLC) kinase and the resultant phosphorylation of the MLC (20, 29). However, this system of thick-filament regulation does not serve as the only mechanism by which smooth muscle contracts, as is evidenced by numerous laboratories showing dissociations between force and MLC phosphorylation levels following initiation of contraction (12, 15, 26, 32). Furthermore, it has been shown that smooth muscle contraction can be initiated in both a calcium- and a MLC phosphorylation-independent manner (11, 33, 37, 38). These findings have led many in the field to study the role of the thin-filament associated proteins in the regulation of smooth muscle contraction, and in particular caldesmon (13).
Caldesmon has generated interest with regards to smooth muscle contraction since its discovery in the early 1980s (34). Caldesmon can be expressed as two isoforms, depending on the primary cell type in which it is expressed. The light (l)-isoform is expressed in both nonmuscle and muscle cells, whereas, in contractile smooth muscle cells, caldesmon is also expressed as the heavy (h)-isoform (24, 40). Caldesmon binds both actin and myosin primarily at its COOH-terminus and NH2-terminus, respectively (13), resulting in the inhibition of actin-activated myosin ATPase activity (4), and, therefore, inhibits force production. Thus caldesmon has gained much attention as a potential protein regulating force production at the level of the contractile apparatus.
Numerous studies have investigated the biochemical and physical chemistry of caldesmon in vitro (13, 17, 23, 24), but there has been little work done indicating a physiologically relevant role(s) for caldesmon as part of the contractile apparatus in vivo. Studies using antagonist peptides made against caldesmon's actin binding domain performed in isolated smooth muscle cells provided some of the first evidence that caldesmon has the ability to inhibit basal levels of contractile tone (21). Our laboratory has shown that downregulating caldesmon using antisense oligodeoxynucleotides results in a tissue with high levels of unstimulated tone (8). These studies, while significant, contribute to a small pool of literature investigating a role for caldesmon in a closer to in vivo situation.
In addition to caldesmon inhibiting basal levels of contractile tone, there have also been theories put forth that caldesmon may have a role in the tonic maintenance of force production. Upon stimulation, a tonically contracting smooth muscle cell generates a peak level of force, which then plateaus, while cytosolic calcium and MLC phosphorylation levels begin to fall to basal levels, termed the latch state (6, 7, 10, 16). Caldesmon has been proposed to play a role in maintaining the high level of force during the latch state (31); however, it has not been fully determined as to whether or not this is one of its physiological roles with regards to contraction. Another area to be investigated with regards to a possible role for caldesmon is the theory of cooperativity, which suggests that cooperative activation of nonphosphorylated cross bridges occurs via phosphorylated cross bridges (36, 39, 39). It has been proposed that caldesmon may regulate the cooperative activation of cross bridges, but there has been little evidence supporting this potential role for caldesmon in vivo.
The goal of this study was to better define the role(s) for caldesmon in intact vascular smooth muscle using short interfering RNA (siRNA) technology. We hypothesized that an h-caldesmon-deficient tissue will result in increased basal levels of tone, in addition to increased physiological contractile properties. Moreover, we were interested in determining whether the relationship between MLC phosphorylation and force would be altered in the h-caldesmon-deficient tissue in such a way as to either support or refute a role for caldesmon in cooperativity between cross bridges during a maintained contraction.
MATERIALS AND METHODS
Swine carotid arteries (SCA) were obtained from a local slaughterhouse and transported to the laboratory in ice-cold physiological salt solution (PSS). PSS contained (in mM) 140 NaCl, 4.7 KCl, 1.2 MgSO4, 1.6 CaCl2, 1.2 Na2HPO4, 0.02 EDTA, and MOPS (pH 7.4). Arteries were cleaned of excess fat and connective tissue and then dissected free of both intimal and adventitial layers, leaving a thin medial layer for experimentation. Strips of the intact carotid media were cut (7 × 0.7 mm) and were stored in PSS (as described above) containing 5 mM d-glucose at 4°C.
Introduction of siRNA and organ culture of vascular strips.
Medial strips of SCA were subjected to a chemical loading protocol originally described by Morgan and Morgan (28) for the introduction of aequorin into intact smooth muscle tissue and similar to that previously published by our laboratory for the introduction of antisense oligodeoxynucleotides (8) and siRNA (33).
The carotid strips were incubated on ice in 1.5-ml Eppendorf tubes (4 strips per 1 ml per tube) with agitation for 90 min in a solution containing the following (in mM): 10 EGTA, 120 KCl, 2 MgCl2, 5 ATP, and 20 HEPES (pH 6.8). The strips were then transferred to a solution containing (in mM) 120 KCl, 2 MgCl2, 5 ATP, and 2 HEPES (pH 6.8) and agitated on ice for 30 min. The strips were then divided into three groups, control (no siRNA), experimental (siRNA against caldesmon), and negative control [siRNA against enhanced green fluorescent protein (eGFP)], and incubated on ice for 90 min while agitated in the same solution (control), or the same solution with siRNA against caldesmon, or siRNA against eGFP. The final concentration of siRNA used was 600 nM, as suggested by the vendor. All strips were then incubated for 30 min on ice with agitation in a similar solution containing the appropriate 600 nM siRNA, and the [MgCl2] (where brackets denote concentration) was increased to 10 mM. Following this incubation, the strips were washed twice with 500 μl of a solution composed of the following (in mM): 140 NaCl, 5 KCl, 10 MgCl2, 5.6 d-glucose, 5 ATP, and 2 HEPES (pH 6.8), and then incubated for 30 min at room temperature with agitation in the same solution with 600 nM siRNA added to the appropriate tubes. At the end of this incubation, all strips were introduced to increasing CaCl2 concentrations every 15 min from 0.001, 0.01, and 0.1, to a final concentration of 1.6 mM. The strips were then soaked in filter-sterilized PSS for 30 min at room temperature.
The strips were placed four strips to a well in a six-well culture plate. Each well was coated with a polymerized silicon elastomer and then filled with 5 ml sterile PSS. The strips were mounted onto the elastomer using 0.2-mm-wide sterilized stainless steel pins. The pins were positioned at the corners of the strips, such that the strips were gently stretched when lifted 2–3 mm above the elastomer. The PSS was then removed by vacuum suction and replaced with 5 ml serum-free medium composed of DMEM/F-12, 100 U/ml penicillin G, 100 μg/ml streptomycin, 35 mg/ml l-ascorbic acid, 200 μg/ml l-glutamine, and 1× insulin-transferrin-selenium. The wells mounted with the siRNA loaded strips also received 600 nM siRNA. The strips were maintained at 37°C with 5% CO2 for a maximum of 10 days after mounting. The medium was changed daily, preceded by a 10-min wash with sterile PSS, stimulation with 110 mM sterile KCl-PSS (equimolar substitution for NaCl) for 10 min, and a second wash in sterile PSS for 10 min and then 5 ml serum-free media overnight. Our laboratory has found that the daily exposure to a KCl-PSS/PSS cycle maintains contractile viability (8, 33).
Immunoblots of h-caldesmon.
Strips of SCA were removed from culture at days 2, 4, 6, 8, and 10 and incubated in PSS warmed to 37°C for 10 min, rinsed in acetone, and were completely air dried. The strips were homogenized in a 1% SDS, 10% glycerol, and 1 mM DTT solution using glass/glass homogenizers. Samples were centrifuged for 5 min, assayed for protein concentration, and subjected to one-dimensional gel electrophoresis on a two-step 7.5 and 14% SDS gel, and transferred to a nitrocellulose membrane at 0.8 A for 2.5 h at 4°C. Following transfer, the membranes were cut at the appropriate molecular marker to be probed separately for h-caldesmon (∼120 kDa) and glyceraldehyde phosphate dehydrogenase (GAPDH) (∼36 kDa). The membranes were blocked in 5% milk-phosphate-buffered solution (PBS) for 60 min at room temperature and then incubated in 1% milk-PBS containing either primary antibody directed at h-caldesmon (1:300) (Biomeda, Foster City, CA) or GAPDH (1:20,000) overnight at 4°C. The membranes were washed twice with PBS each for 10 min, followed by a 10-min wash in 0.05% Tween-PBS, and then incubated in 1% milk-PBS containing a sheep anti-mouse secondary antibody (1:10,000) for 90 min at room temperature. The washes were repeated, and the membranes were incubated in an enhanced chemiluminescence (ECL) solution for 1 min and developed. Based on preliminary studies, all ECL developed films were in the linear range for quantitative densitometric analysis.
Immunoblots of h- and l-caldesmon.
Samples were subjected to SDS-PAGE electrophoresis and transferred to nitrocellulose membranes, as described above. Following transfer, the membranes were incubated in primary antibody that recognizes both h- and l-caldesmon (∼120 and 90 kDa, respectively) and α-actin (∼42 kDa). The membranes were blocked in Odyssey blocking buffer for 50 min at room temperature, followed by incubation in a 1:1 mixture of Odyssey blocking buffer and 0.2% Tween-PBS containing primary antibody directed at h- and l-caldesmon (1:10,000) and α-actin (1:2,000,000) overnight at 4°C. The membranes were washed five times with 0.1% Tween-PBS each for 5 min, followed by a 5-min wash in PBS, and then incubated in a 1:1 mixture of Odyssey blocking buffer and 0.2% Tween-PBS containing a fluorescent anti-mouse secondary antibody (1:10,000) for 45 min at room temperature. The washes were repeated, and the membranes were developed using the Odyssey imager. Any blot containing saturated bands was not included in the analysis.
Immunohistochemistry for h-caldesmon.
SCA strips were cultured as described above and then removed, placed in Histochoice, and sent to AML Laboratories (Rosedale, MD) for paraffin embedding and sectioning. Paraffin was removed from the cross sections (5 μm) by heating to 70°C and then using xylene substitute twice for 10 min each. Slides were then put into descending grades of alcohol (100, 70, and 30%) for 10 min each step and then in 1× PBS for three washes (10 min each wash). After washing, the section slices were blocked with a 1% BSA-PBS solution for 1 h at room temperature, followed by PBS containing primary antibody directed against h-caldesmon (Biomeda, Foster City, CA) (1:300) overnight at 4°C. The slides were washed using PBS three times, 10 min each, followed by incubation in PBS containing secondary antibody anti-mouse IgG-Cy3 (1:400, Sigma Chemical, St. Louis, MO) for 1 h at room temperature. Slides were washed in PBS three times, 10 min each. A drop of mounting medium was applied to each slide before adding a coverslip. Sections were viewed under a Nikon (Melville, NY) Eclipse E800 fluorescent microscope, and images were captured using a RT Slider SPOT camera (Diagnostic Instruments, Sterling Heights, MI) and Image-Pro Plus software from Media Cybernetics (Bethesda, MD). A negative control, in which only the secondary antibody was added, was performed for all samples.
Electrophoresis of myosin isoforms.
Aliquots (15 μg) of total protein from control and SCA strips cultured for 8 days were subjected to SDS-PAGE gel electrophoresis (3% stacking gel, 5% separating gel). Gels were stained with Coomassie blue to visualize proteins. The gels were scanned using a Bio-Rad GS800 densitometer to quantify electrophoretic spots corresponding to the SM1 and SM2 isoforms of myosin. Results are presented as the ratio of the densitometric scan of SM1 to that of SM2.
Tissue strips were removed from culture and allowed to soak in PSS warmed to 37°C for 10 min. The strips were suspended between a Grass FT.03 force transducer and a stationary clip in water-jacketed organ baths. The strips were stretched to 5 g of force and equilibrated in 15 ml PSS aerated with 100% O2 and warmed to 37°C for 90 min. The strips were subjected to several exposures of 110 mM KCl-PSS and then allowed to relax. The length of the tissues was adjusted to be just below slack length. The length of each strip was measured, then the tissues were maximally contracted with 110 mM KCl-PSS for 5 min, and active force at the peak of contraction was recorded. The strips were allowed to relax in PSS and then slightly stretched. The new length was recorded, and the tissue was maximally contracted again. This process was repeated until active force decreased ∼50% from that observed to be peak active force. The length and force data points were normalized to Lo and Fo, respectively, where Lo was the length at which maximal active force was achieved, and Fo was the maximal active force at Lo. Passive forces were also measured at each tissue length.
Measurement of isotonic shortening velocity.
Tissue strips were removed from culture and mounted between a clip attached to a micrometer on one end for control of initial muscle length, and by an aluminum foil tube connected to a Cambridge Technology 300H servo lever interfaced to a Linux-based microcomputer equipped with custom software to control length and force, as well as fit the collected data for analysis of shortening. Strips were incubated in 60 ml PSS at 37°C, aerated with 100% O2, stretched to ∼5 g of force, and allowed to equilibrate for ∼90 min. After equilibration, the strips were stimulated with 110 mM KCl-PSS several times until a stable force was achieved. Velocity of shortening was measured at various times during a contraction in response to 10 μM histamine. After the addition of histamine, the strips were subjected to a series of isotonic releases equal to 5, 10, 15, and 20% of the force at the time of release. Isotonic releases were completed within 1.3 ms. These releases were performed 1, 5, 10, and 20 min following initiation of contraction. At the 1-min time point, only one release was performed during a single contraction. The other three releases were performed on separate contractions of similar magnitude in the same arterial strip. The change in length during each force clamp was stored on the computer, and the change in length vs. time at 100 ms after the release was recorded as muscle lengths per second. To measure shortening velocity in the unstimulated strips, organ cultured strips were mounted and equilibrated as described above, stretched to approximately the same force as was generated in response to 10 μM histamine, and allowed to stress relax. The tissues were repeatedly stretched until they maintained the required resting force. The tissues were then subjected to a rapid isotonic release to 5% of the resting force at the time of release. The change in length during the force clamp was stored, and the change in length vs. time at 100 ms after release was recorded as muscle lengths per second.
Isometric force development.
Tissue strips were removed from culture and allowed to soak in PSS warmed to 37°C for 10 min. The strips were suspended between a Grass FT.03 force transducer and a stationary clip in water-jacketed organ baths. The strips were stretched to 5 g of force, allowed to equilibrate in 15 ml PSS aerated with 100% O2, and warmed to 37°C for 90 min. A passive force of ∼2 g was applied to all strips, followed by stimulation with 110 mM KCl-PSS, relaxed with PSS, and stimulated again to generate force. The strips were subjected to the cumulative addition of histamine (1–30 μM) for a total of 20 min. After complete relaxation, the strips were dried and subjected to Western blot analysis for h-caldesmon content.
Triton X-100 skinning and MLC phosphorylation.
Tissue strips were removed from culture and mounted for isometric force recording, as described above. The strips were stimulated with 110 mM KCl and allowed to completely relax and stimulated again with 110 mM KCl for 5 min. The strips were washed in a Ca2+-free PSS solution containing 1 mM EGTA for 10 min, after which 10 μM histamine was added for 30 min to deplete intracellular calcium. The strips were then incubated for 60 min in a 0.5% Triton X-100 skinning solution containing (in mM) 5 EGTA, 20 imidazole (pH 6.8), 50 potassium actetate, 1 DTT, and 150 sucrose. All strips were then incubated for 15 min in a high-EGTA relaxing solution containing (in mM) 5 EGTA, 20 imidazole (pH 6.8), 50 potassium actetate, 1 DTT, 6 MgCl2, and 6 ATP. The strips were then incubated for 15 min in a low-EGTA relaxing solution containing (in mM) 0.1 EGTA, 20 imidazole (pH 6.8), 50 potassium actetate, 1 DTT, 6 MgCl2, and 6 ATP. The strips were then stimulated for 20 min with a Ca2+ contracting solution, using the following concentrations (in μM): 1.0, 3.0, 5.0, 7.0, or 10.0, containing (in mM) 5 EGTA, 20 imidazole (pH 6.8), 50 potassium actetate, 1 DTT, 1 Mg2+, and 4 mM ATP. Tissue strips were then rapidly frozen in a dry ice/acetone slurry containing 6% trichloroacetic acid and 10 mM DTT, slowly thawed to room temperature, rinsed in acetone, air dried, and subjected to homogenization in a 1% SDS, 10% glycerol, and 1 mM DTT solution using glass/glass homogenizers. Samples were centrifuged for 5 min, subjected to two-dimensional gel electrophoresis, and transferred to nitrocellulose membranes for quantification of MLC phosphorylation, as previously described (27). Transferred proteins were visualized with Colloidal Gold and digitized by densitometric analysis using a Bio-Rad GS-800 quantitative densitometer. MLC phosphorylation levels were calculated by dividing the densitometric analysis of the spot corresponding to the phosphorylated MLC by the sum of the densitometric analyses of the spots corresponding to the phosphorylated and unphosphorylated MLC. Values are presented as moles Pi per mole MLC.
Materials and statistics.
siRNA against caldesmon was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Primary antibody for h-caldesmon for use with the ECL developed Western blots was purchased from Biomeda (Foster City, CA). Goat anti-rabbit secondary antibody was purchased from Upstate Technology (Billerica, MA). Fluorescent anti-mouse secondary antibody used with the Odyssey imager was purchased from LI-COR (Lincoln, NE). Primary antibodies for h- and l-caldesmon and α-actin for use with the Odyssey imager, anti-mouse secondary fluorescently conjugated antibody, ATP, creatine phosphate, histamine, EGTA, and Triton X-100, were obtained from Sigma Chemical. Culture media was obtained from Mediatech (Manassas, VA), and media antibiotics and reagents were obtained from Invitrogen (Carlsbad, CA). Xylene substitute was purchased from Shandon (Pittsburgh, PA). Mounting medium was purchased from Vector Laboratories (Burlingame, CA). All protein assays and electrophoretic and Western blotting chemicals and materials were obtained from Bio-Rad (Richmond, CA). Sheep anti-mouse secondary antibody, Western blotting detecting reagents, hyperfilm, and Colloidal Gold were all purchased from Amersham Biosciences (Piscataway, NJ). GAPDH primary antibody was purchased from Advanced Immunochemical (Long Beach, CA). All other reagents were obtained from Fisher Scientific (Pittsburgh, PA) and were analytical grade or better.
All values shown are means ± SE with “n” representing the number of vascular preparations, each taken from a different artery. Statistical significance was determined using the Student's t-test for unpaired values. A value of P < 0.05 was taken as significant.
siRNA-mediated downregulation of h-caldesmon.
Carotid arterial strips were dissected and subjected to the chemical loading protocol, as described in the materials and methods section. Tissue strips were either chemically loaded without siRNA (control) or with 600 nM siRNA targeted against h-caldesmon and then cultured for 2, 4, 6, 8, and 10 days and processed for measurement of h-caldesmon content. Figure 1A shows representative immunoblots of carotid tissues subjected to organ culture in control or siRNA against caldesmon. Several blots of the type shown in Fig. 1A were scanned for h-caldesmon content and were normalized to GAPDH as a loading control in both the control and siRNA-treated tissues at each day of culture. GAPDH levels were not affected by organ culture. The siRNA-induced decrease in h-caldesmon, compared with the same number of days of vehicle organ culture, is presented as a percentage of h-caldesmon content in control, noncultured tissues in Fig. 1B. h-Caldesmon levels in control organ cultured tissues did not change during the 10-day experimental period. After 8 days of culture in the presence of siRNA, h-caldesmon levels were significantly decreased by ∼60% of the levels present in control, noncultured strips at day 0. h-Caldesmon content rebounded to levels approximating noncultured tissues by day 10, consistent with our laboratory's previous studies showing that casein kinase II also rebounded to control levels with extended culture (33). Whether this is due to recovery from siRNA inhibition or cellular degradation of the siRNA is not known. All subsequent studies were performed on strips cultured for 8 days in the presence of the siRNA against h-caldesmon.
To further support the Western blot analysis of the siRNA-mediated downregulation of h-caldesmon shown in Fig. 1, immunohistochemistry was performed on tissues removed from culture on day 8. More importantly, we were interested in determining whether siRNA decreased h-caldesmon content uniformly across the entire tissue or resulted in a tissue with very low levels of h-caldesmon on the perimeter of the tissue and normal levels in the central core of the tissue. The same antibody that was used to probe for h-caldesmon in Western blots was used for immunohistochemical imaging. Representative immunohistochemical images from tissues removed from culture on day 8 are shown in Fig. 2. There is intense visible fluorescent red staining for h-caldesmon in the control day 8 organ cultured tissue strips (Fig. 2A) compared with the visibly less intense fluorescent red staining in the siRNA-treated tissue strips (Fig. 2B). These results further demonstrate that introduction of siRNA into the cultured vascular strips mediated a significant downregulation of h-caldesmon. These data also suggest that the siRNA is evenly diffused throughout the tissue, resulting in a uniform decrease in h-caldesmon expression across the entire thickness of the vascular strip.
Many antibodies used to probe for caldesmon are nonspecific and are able to detect both the l- and h-caldesmon isoforms. To test the specificity of the h-caldesmon antibody used in the Western blot and immunohistochemical imaging, a negative control with 3T3 cells and fibroblast cells was run with the tissue samples and subjected to Western blot analysis using the h-caldesmon antibody referenced above and an antibody considered to be nonspecific and reacts with both h- and l-caldesmon. Both controls were positive for l-caldesmon using the nonspecific caldesmon antibody, but negative using the h-caldesmon-specific antibody (data not shown).
Smooth muscle cells in culture have the ability to phenotypically modulate from a contractile cell to a synthetic cell. This switch is accompanied by changes in smooth muscle cell-specific proteins, as well as changes in the regulation of the contractile proteins, including an alteration in caldesmon isoform expression from h- to l-caldesmon (25, 30, 35). To ensure that culturing or siRNA introduction had only a minimal if any, effect on the phenotype of the tissues, l-caldesmon levels were measured. There were no increases in l-caldesmon expression as a result of culture or introduction of siRNA against h-caldesmon after 8 days of culture, as shown by a representative blot in Fig. 3A and by quantitation of several such blots in Fig. 3B, left.
We used siRNA against eGFP as a negative control for potential nonspecific effects of the introduction of a siRNA. siRNA against eGFP has been successfully used before as a negative control in muscle cells (42). Figure 3A shows a representative blot of SCA tissues cultured for 8 days in the presence of siRNA against eGFP. The quantitation of several such blots shows that siRNA against eGFP had no effect on h-caldesmon levels (Fig. 3B right).
As discussed above, exposing smooth muscle to culture can alter the phenotype of the muscle cells. One of the primary changes that can occur is a change in the ratio of myosin isoforms. We quantified the effect of 8 days of organ culture on the ratio of SM1/SM2 isoforms of myosin, and these results are shown in Fig. 4. Figure 4A shows a representative Coommassie blue stained gel, while Fig. 4B shows the quantitation of several such gels. Eight days of culture had no effect on the ratio of SM1/SM2 isoforms of myosin.
Length/tension relationship in tissues subjected to organ culture.
SCA tissues exposed to organ culture for 8 days in the presence of vehicle or siRNA against h-caldesmon were subjected to experiments to measure their length/tension relationships. This was performed to determine whether knockdown of h-caldesmon had any significant effect on the relationship between tissue length and force. The results are shown in Fig. 5. There were no significant differences in the length/tension relationship of tissues subjected to organ culture for 8 days in the presence of siRNA against h-caldesmon compared with control tissues cultured for 8 days. There were also no significant differences in the length/passive force relationships between the two cultured groups of tissues (data not shown).
Effect of caldesmon on isotonic shortening velocity.
Caldesmon is believed to be important in maintaining low levels of basal tone, as evidenced by the finding that downregulation of caldesmon levels results in a constitutively active tissue (8). We were interested in more precisely studying this potentially important role of h-caldesmon using siRNA technology and a more quantitative approach. We quantified isotonic shortening velocities in tissues subjected to siRNA knockdown of h-caldesmon levels.
Tissue strips were removed from culture on day 8 following introduction of siRNA and mounted for isotonic force recording. The tissues were stimulated with 110 mM KCl, relaxed and stimulated again with 10 μM histamine, followed by complete relaxation and several rinses with PSS. The tissues were stretched to the amount of force generated by the 10 μM histamine-induced contraction and allowed to stress relax. The stretch/stress relaxation cycle was repeated until the tissue strips maintained the unstimulated stretched force. The strips were then subjected to a quick release to 5% of the unstimulated force present at the time of release. The results of these experiments are shown in Fig. 6. Tissues not cultured and tissues cultured but not exposed to siRNA exhibited very low levels of isotonic shortening velocity following a 5% release from the unstimulated stretched force. In contrast, the tissues with knockdown of h-caldesmon exhibited significant levels of shortening velocity, suggesting the presence of active cross-bridge cycling in the unstimulated tissues.
To measure isotonic shortening velocity in the stimulated tissues, carotid arterial strips were mounted after 8 days of culture, as described above, stimulated with 110 mM KCl, relaxed completely, followed by stimulation with 10 μM histamine. At 1, 5, 10, and 20 min following initiation of the histamine-induced contraction, the strips were released to 5, 10, 15, and 20% of the stimulated force at the time of the release. As discussed in materials and methods, only one release was performed per contraction at the 1-min time point. As shown in Fig. 7, h-caldesmon-depleted tissues had significantly slower shortening velocities (expressed as muscle lengths/s) at most time points of the contraction compared with control tissues.
Determine the effect of h-caldesmon downregulation on force.
To determine the role of caldesmon on force, we measured the response of control and h-caldesmon-depleted tissues to histamine stimulation (Fig. 8). Tissues were cultured for 8 days, mounted for isometric force recording, and stimulated by the cumulative addition of histamine to produce agonist-induced contractions. The findings in Fig. 8 demonstrate that the h-caldesmon-deficient tissues tended to produce less stress, both at peak levels (panel A) and steady-state levels (panel B, determined at 20 min after initiation of contraction) in response to histamine compared with control tissues. While the decreased levels of stress production were not significantly different between the h-caldesmon-deficient and control tissues, the trend for lower stimulated force production was consistent over the entire [histamine] range, as well as with a previous study from our group that showed significant decreases in histamine-induced stress in a caldesmon knockdown tissue (8).
Most, if not all, tonic as well as phasic smooth muscles exhibit a state of high force supported by low levels of MLC phosphorylation, the hypothesized latch state (6, 7, 10, 16). Caldesmon has been suggested to be important in the maintenance of force without proportional MLC phosphorylation. If this potential role for caldesmon is plausible, then force should not be maintained in the caldesmon-deficient tissues. Steady-state levels of stress in the h-caldesmon-deficient tissues were not significantly different than peak levels of stress (Fig. 8B). This would suggest that h-caldesmon is not involved in the maintenance of force during the hypothesized latch state.
Determine the effect of h-caldesmon downregulation on the cooperativity of smooth muscle contraction.
Work from the Siegman and Butler laboratories (39) suggested that phosphorylated cross bridges may cooperatively activate unphosphorylated cross bridges during the tonic maintenance of force. To determine whether caldesmon plays a role in the cooperative activation of unphosphorylated cross bridges, we measured the calcium/MLC phosphorylation and MLC phosphorylation/force relationships in control and h-caldesmon-depleted tissues. Tissue strips were removed from culture and mounted for isometric force recording. Steady-state levels of stress and MLC phosphorylation were measured in Triton X-100 skinned preparation in response to stimulation with 0, 3, 5, or 7 μM Ca2+ for 20 min. Basal values of MLC phosphorylation are slightly higher than those that have been previously reported, suggesting the possibility that the combination of organ culture and Triton X-100 skinning may decrease MLC phosphatase levels. Figure 9 shows the results of these experiments. Knockdown of h-caldesmon had no significant effect on either the MLC phosphorylation dependence on Ca2+, or the stress dependence on MLC phosphorylation. The lack of change in the latter relationship would suggest that h-caldesmon is not involved in the cooperative activation of unphosphorylated cross bridges.
One hallmark of smooth muscle contraction is the high levels of force that can be supported without proportional levels of MLC phosphorylation or calcium (6, 7, 16, 29). Although this phenomenon has been shown in many tissues from many species, the underlying mechanism(s) responsible is not known. Due to the fact that force can be maintained without proportional levels or, in some cases, the absence of MLC phosphorylation, attention has been focused on thin-filament based proteins (13, 22). The thin-filament protein, caldesmon, has received the most attention; however, few studies have provided results suggesting a physiologically relevant role (3, 31). Our laboratory has previously shown that knockdown of caldesmon using antisense oligonucleotides in conjunction with organ culture of the swine carotid medial strip results in a tissue that exhibits active cross-bridge cycling in the unstimulated state (8). Unfortunately, this finding has not been reproduced or corroborated by other investigators. Therefore, one of the goals of this study was to extend our previous findings using enhanced techniques. Specifically, we were interested in providing new mechanistic information on how h-caldesmon may potentially regulate or modulate vascular smooth muscle contraction to elucidate caldesmon's role(s) in a closer to in vivo situation.
The results of our present study provide, we believe, novel and potentially important results concerning h-caldesmon and the regulation of vascular smooth muscle. First, we have corroborated our original finding that knockdown of h-caldesmon produces a tissue with constitutively active cross-bridge cycling. Of particular importance is that we confirmed our earlier results using a newer and more specific protein knockdown technique and a more refined mechanical technique. Our laboratory first demonstrated that siRNA could be used to knock down specific proteins in smooth muscle tissues in culture in a study on casein kinase II (33). Moreover, our laboratory has previously shown that introduction of siRNA into swine carotid vascular strips has little to no effect on viability and does not increase interferon levels (33). Our techniques were later shown to work in rat cerebral arteries (5). In this present study, we extend our use of siRNAs to show that the smooth muscle-specific protein h-caldesmon can be specifically knocked down in a representative vascular preparation. This knockdown of h-caldesmon occurs without either a knockdown of, or a compensatory increase in, the nonsmooth muscle-specific isoform, l-caldesmon. One concern in this approach is if introduction of siRNA results in a near complete knockdown in cells in the edges of the tissue, while cells in the core of the tissue have near normal protein levels. Our immunostained tissue sections (Fig. 2) show that siRNA produced a reasonably homogenous knockdown throughout the tissue. It is possible that there is more than one pool of caldesmon, and our studies only knocked down one specific pool. For example, if one pool of caldesmon has a faster turnover rate, this may explain why we could not achieve closer to 100% knockdown. Although not conclusive, our immunostained tissue sections suggest uniform knockdown rather than a decrease in a select pool. Therefore, we believe these results emphasize the significance of using siRNA technology in tissue to better understand contractile mechanisms in a closer to in vivo setting, eliminating the need to use cultured cells, which typically modulate their phenotype, making contractile studies difficult.
siRNA knockdown of h-caldesmon produced a vascular tissue that exhibited active cross-bridge cycling in the resting unstimulated state. The results from the present study corroborate, extend, and supplement our laboratory's previous findings (8) by quantifying the rate of cross-bridge cycling in the unstimulated tissue. Our earlier study only demonstrated that a quick release of an unstimulated stretched tissue, containing 40% normal h-caldesmon content, redeveloped force, indicative of active cross-bridge cycling. In the present study, we show that unstimulated h-caldesmon knockdown cultured tissues shortened following a quick release to a 5% force clamp with a velocity of ∼0.08 muscle lengths/s, ∼65% of the velocity of histamine stimulated caldesmon knockdown tissues (Figs. 6 and 7). Our laboratory previously suggested the possibility that h-caldesmon inhibited a pool of unphosphorylated but active myosin cross bridges (8). However, with the knowledge that isotonic shortening velocity during a 5% force clamp will underestimate the maximal velocity of shortening, then the velocities of the unstimulated and stimulated tissues may not be that dissimilar. This assumption, coupled with the high basal MLC phosphorylation levels (Fig. 6), supports the hypothesis that h-caldesmon may act as a brake on phosphorylated active myosin heads. This would be in agreement with a similar hypothesis proposed previously (9) and is the simplest explanation of our results.
Caldesmon may have other regulatory roles in smooth muscle contraction, in addition to potentially acting as a brake against basally phosphorylated and, therefore, active myosin cross bridges. Two roles that have been proposed for caldesmon are in the maintenance of force supported by un- or dephosphorylated cross bridges (6, 7, 10, 16) and in the cooperative activation of unphosphorylated cross-bridge interaction with actin by phosphorylated cross bridges (36, 39, 41).
As discussed previously, one of the hallmarks of smooth muscle contraction is the ability to maintain high levels of force during conditions of low levels of MLC phosphorylation and calcium (6, 7, 10, 16). One reasonably popular hypothesis to account for the high force maintenance is cross-linking of the thick and thin filaments by caldesmon. The COOH-terminal of caldesmon has an actin-binding domain, while the NH2-terminal contains a myosin binding domain (13). Maintenance of high levels of force has been proposed to be the result of caldesmon “locking” the contractile filaments in a force-bearing state (4). If this hypothesis is correct, then one would expect that the high-force state in the absence of proportional levels of MLC phosphorylation and calcium would be absent in a h-caldesmon knockdown smooth muscle tissue. However, our data suggest that h-caldesmon knockdown had no significant effect on the magnitude or time course of steady-state maintained force in the histamine-stimulated swine carotid medial tissue (Fig. 8B). This lack of effect was not the result of a compensatory increase in l-caldesmon (Fig. 3) or a selective loss of h-caldesmon only in the cells on the periphery of the tissue (Fig. 2). It is possible that the remaining ∼40% h-caldesmon in the knockdown tissue was sufficient to maintain force. However, it would seem unlikely that a 60% decrease in protein content would have no effect if h-caldesmon was important in the maintenance of force.
A second thin-filament-based protein, calponin, has also been suggested to be involved in the maintenance of force. Although calponin levels were not overexpressed in the h-caldesmon knockdown tissue (data not shown), it remains a possibility that calponin is important in the regulation of maintained force. Conversely, calponin may be able to assume the role played by caldesmon in tissues with depressed h-caldesmon expression levels. This hypothesis is based on the findings that h-caldesmon and calponin compete for the same binding sites on actin (18). If h-caldesmon levels are decreased and calponin is free to bind to actin and serve the same or similar role as h-caldesmon, then h-caldesmon depletion would be compensated for by calponin, resulting in no net difference in force maintenance. Whether this potential functional role for calponin can occur without the h-caldesmon myosin domain remains unknown.
The cooperative activation of unphosphorylated cross bridges by phosphorylated cross bridges has been proposed as a mechanism to explain force development without proportional levels of MLC phosphorylation (36, 39, 41). It is possible that caldesmon transmits the cooperativity signal along the thin filament. Conversely, it has been suggested that caldesmon inhibits cooperativity (1, 19). If caldesmon played a positive role in cooperativity, one would expect a shift to the right in the MLC phosphorylation vs. force relationship. If caldesmon played an inhibitory role in cooperativity, one would expect a shift to the left in the same relationship. Our results show no shift in the curve in either direction following knockdown of h-caldesmon (Fig. 9B). Therefore, our results do not support a role for h-caldesmon in inhibition or transmission of cooperativity. Our results are not, however, entirely consistent with reports in the literature. Ansari et al. (2) recently presented an elegant study demonstrating that caldesmon cooperatively switches actin to an “off” state, while Ca2+-calmodulin-caldesmon switch actin to an “on” state. Thus activation of the contractile elements can be enhanced in a caldesmon-dependent manner at constant levels of MLC phosphorylation. Consistent with this is earlier work from Haeberle (14), who presented results showing that thin filaments “turned-on” by NEM-S1 exhibited greater force and velocity in an in vitro motility assay using unphosphorylated myosin compared with thin filaments not “turned-on”. One possibility to explain these differences is that caldesmon binding is required for the cooperative switching of actin to an “off” or “on” state, as shown by Ansari et al. (2). In our knockdown preparation, caldesmon is not present and, therefore, not able to interact with actin. Further studies are required to determine whether this hypothesis is correct.
Histamine induced force development and maintenance tended to be depressed in the h-caldesmon knockdown tissue compared with control tissues (Fig. 5, A and B). Force responses in our laboratory's previous report (8) were significantly depressed in the caldesmon knockdown tissues compared with control. We do not believe this is an inhibitory effect of h-caldesmon knockdown per se. Any given muscle has a finite level of stress that can be maximally developed. The h-caldesmon knockdown tissues exhibit active cross-bridge cycling in the unstimulated state; therefore, the muscle cells must have some degree of contractile activity and developed stress in the basal resting condition. One could think of the knockdown tissues as precontracted before stimulation due to the removal of the brake against phosphorylated cross bridges at rest. The addition of histamine will produce a full contractile response in the control tissues, but can only produce the difference between the maximal force attainable by the carotid medial tissue and the “precontracted” level of force as a result of h-caldesmon knockdown. If our reasoning is correct, then we could assume that h-caldesmon inhibits ∼18–20% of the total maximal force that the swine carotid media is able to develop. Therefore, in the absence of h-caldesmon, a blood vessel could conceivably contract up to 20% of its maximal force-generating capacity, with the resultant significant increase in resistance and, therefore, blood pressure.
In summary, we have provided additional evidence demonstrating that our technique (33) using siRNA, in conjunction with organ culture of vascular smooth muscle, to produce a viable tissue with the knockdown of a specific protein is indeed an effective tool to study the regulation of smooth muscle. Based on our results, h-caldesmon does not appear to be involved in maintenance of force supported by low levels of MLC phosphorylation and may not be involved in the cooperative activation of unphosphorylated cross bridges by phosphorylated cross bridges, although h-caldesmon may be important in thin-filament activation. On the other hand, knockdown of h-caldesmon by siRNA produces a vascular tissue with constitutive activity. The velocity of shortening of the constitutively active tissue and the high basal values of MLC phosphorylation suggest that h-caldesmon in vivo acts as a brake against unwanted contractions due to basally phosphorylated myosin. It is also possible that phosphorylation of h-caldesmon alone in the resting state may be a mechanism to produce smaller increases in force without receptor-dependent stimulation or increases in cytosolic calcium. Disinhibition of h-caldesmon by phosphorylation would then allow force to be developed by activated myosin in the resting state.
This study was supported, in part, by National Institutes of Health Grants HL 37956 and DK 69898, and an O'Brien Urology Center Grant to the University of Pennsylvania. E. Smolock was supported by a Predoctoral Fellowship from the Mid-Atlantic Chapter of the American Heart Association.
- Copyright © 2009 the American Physiological Society