Caveolin-1 (Cav-1) is essential for the morphology of membrane caveolae and exerts a negative influence on a number of signaling systems, including nitric oxide (NO) production and activity of the MAP kinase cascade. In the vascular system, ablation of caveolin-1 may thus be expected to cause arterial dilatation and increased vessel wall mass (remodeling). This was tested in Cav-1 knockout (KO) mice by a detailed morphometric and functional analysis of mesenteric resistance arteries, shown to lack caveolae. Quantitative morphometry revealed increased media thickness and media-to-lumen ratio in KO. Pressure-induced myogenic tone and flow-induced dilatation were decreased in KO arteries, but both were increased toward wild-type (WT) levels following NO synthase (NOS) inhibition. Isometric force recordings following NOS inhibition showed rightward shifts of passive and active length-force relationships in KO, and the force response to α1-adrenergic stimulation was increased. In contrast, media thickness and force response of the aorta were unaltered in KO vs. WT, whereas lumen diameter was increased. Mean arterial blood pressure during isoflurane anesthesia was not different in KO vs. WT, but greater fluctuation in blood pressure over time was noted. Following NOS inhibition, fluctuations disappeared and pressure increased twice as much in KO (38 ± 6%) compared with WT (17 ± 3%). Tracer-dilution experiments showed increased plasma volume in KO. We conclude that NO affects blood pressure more in Cav-1 KO than in WT mice and that restructuring of resistance vessels and an increased responsiveness to adrenergic stimulation compensate for a decreased tone in Cav-1 KO mice.
caveolae are 50- to 100-nm large membrane invaginations in which cholesterol and sphingolipids are enriched. Two caveolin protein homologs (Cav-1 and -3) are necessary for the formation of caveolae and act as scaffolds for many signaling molecules (4, 21, 28). A prototypical example of a caveolae-associated signal-transduction molecule is endothelial nitric oxide (NO) synthase (eNOS), which is targeted to caveolae via acylation (29). Cav-1 exerts a negative regulatory influence on eNOS activity, and this is of relevance for NO production (18). Accordingly, Cav-1 ablation, which blocks formation of caveolae in the vasculature, results in increased NO generation and vasodilatation ex vivo (5, 23, 32, 35). Despite increased NO release, a threefold increase in cGMP in smooth muscle cells (5), a major reduction of myogenic tone (1, 5, 7), and the development of heart failure (4, 32, 35), mean arterial blood pressure is largely normal in adult Cav-1 knockout (KO) mice (26).
Blood pressure is essential for homeostasis, and it is expected that a reduction in myogenic tone in KO arteries will be compensated, possibly via the baroreceptor reflex. This would restore blood pressure via increased sympathetic activity and thus neurogenic tone. No major difference in heart rate was, however, found in anesthetized KO relative to wild-type (WT) mice (26). On the other hand, force in response to α1-adrenergic stimulation is increased in denuded femoral arteries from KO mice compared with WT controls (27), suggesting compensation at the level of the vascular wall. Increased contraction in response to noradrenaline was similarly found in the saphenous artery (20). An increased contractile response to α1-adrenergic stimulation would counteract the blood pressure-lowering effect of NO in the vasculature. In addition, hypertrophic remodeling without a matching increase in lumen diameter would amplify the effect of a given vasoconstrictor's influence on arterial resistance (8).
Shear stress by blood flow has been proposed to activate both eNOS activity and the Ras-Raf-MAP kinase pathway within caveolae in the endothelium (24, 25). Accordingly, a recent study (33) demonstrated that carotid arteries from KO mice show impaired flow-mediated dilatation. Such an effect could potentially contribute to altered regulation of vascular resistance in Cav-1 KO mice. However, it should be noted that vascular resistance is also influenced by endothelium-derived hyperpolarizing factor, distinct from NO and prostacyclin (3). Thus the role of flow-dependent vasodilatation in KO mice needs to be studied also in resistance vessels. Finally, plasma volume is an important determinant of venous return to the heart and is an effector mechanism for renin-angiotensin system activation. Plasma volume is known to be expanded in heart failure (16) and may therefore be increased in KO mice, of possible importance for the control of blood pressure.
The aim of the present study was to test the following hypotheses in Cav-1 KO mice: 1) that small mesenteric arteries (SMAs) are remodeled and have increased α1-adrenergic contractility; 2) that flow-mediated dilatation is impaired; and 3) that the plasma volume is expanded. If so, a few candidate mechanisms contributing to blood pressure normalization in the setting of a large constitutive NO production in these mice would have been identified.
MATERIALS AND METHODS
Cav-1 KO mice.
Cav-1 KO mice on the C57BL/6 background obtained from the Jackson Laboratory (Bar Harbor, ME) were further backcrossed on the same background and were genotyped as described by Razani et al. (23). WT littermates or purchased C57BL/6 mice (Möllegard, Copenhagen, Denmark), matched for age and sex, were used as controls (all are referred to as WT); 10- to 16-wk-old mice of both sexes weighing between 20 and 30 g were used. Mice had free access to standard chow and water. For in vitro experiments, mice were euthanized by cervical dislocation. All experiments conformed to the American Physiological Society's “Guiding Principles in the Care and Use of Animals” and were approved by the local animal ethics committee.
Preparation of vessel segments.
The intestine with the mesentery attached, aorta, and portal vein were removed and placed in cold (4°C) HEPES-buffered Krebs solution (composition in mM: 135.5 NaCl, 5.9 KCl, 1.2 MgCl2, 11.6 glucose, 11.6 HEPES, pH 7.4), prepared by using Milli-Q water and containing 2.5 mM Ca2+ as indicated. First-generation SMAs, aortas, and portal veins were carefully freed from adhering tissue under a microscope.
Isometric force recording.
SMA and aorta segments (1.5 mm long) were mounted in a myograph (610M; Danish MyoTechnology, Aarhus, Denmark), as previously described (2). In these experiments, all vessels were denuded of endothelium (6), and 300 μM l-NAME was included, which eliminated relaxation in response to 10 μM acetylcholine, because the objective was to determine the effects of remodeling on force generation irrespective of endothelial NO production. The exact length of each segment was determined by using a microscope with an ocular scale, and force was expressed relative to length. After determination of the passive force for optimal active force development (1.7 mN/mm in both KO and WT, see ⇓⇓⇓⇓Fig. 5 and Flow-induced dilatation and combined tone), SMAs were routinely stretched to 2.5 mN (=1.7 mN/mm). The same basal force was applied to the aorta, because pilot experiments had shown that this was close to L0, the circumference resulting in the greatest active (total minus passive) force in each preparation. Circumference-tension relationships were generated by increasing wire distance in a stepwise fashion. At each wire distance, preparations were allowed to equilibrate for 5 min. High-K+ solution, obtained by exchanging 60 mM NaCl for KCl, was then applied for 7 min. Depolarization was used, rather than noradrenaline, because it generates highly reproducible contraction for repeated challenges. At the end of each cycle, passive force was obtained by relaxing the preparations for 10 min in Ca2+-free solution, which was shown in control experiments to give complete relaxation. Thereafter, wire distance was increased and the cycle was repeated.
Pressure myograph recording.
SMAs with intact endothelium were mounted on glass cannulae (diameter = 100 μm) in a pressure myograph chamber (Living Systems Instrumentation, Burlington, VT) and were secured with silk sutures. The myograph chamber was placed on a Nikon Diaphot 200 inverted microscope equipped with a charge-coupled device (CCD) camera. VediView 1.2 software (Danish MyoTechnology) was used to monitor lumen and vessel diameter and was also used to determine wall thickness and cross-sectional area after full dilatation. The edges chosen by the software were examined critically in all experiments. The vessel was superfused with HEPES-buffered Krebs solution gassed with O2, and temperature and pH (7.4) were monitored. Intraluminal pressure was applied by gravity via the cannula from a solution reservoir mounted at an adjustable height. Pressure was monitored by two pressure transducers mounted at the inflow and outflow line. Flow was applied by using a peristaltic pump, and intraluminal pressure was maintained during this procedure. An expandable rubber tube was mounted at the inflow line to reduce pulsations from the pump.
Before experimentation, intraluminal pressure was gradually increased to 95 mmHg. At this pressure, the vessel was stretched manually in the longitudinal direction to 115% of its unstretched length. Preparations were subsequently equilibrated at 36°C during 1 h at 45 mmHg. Arteries that showed signs of leakage, identified by using a drop chamber mounted on the inflow line, were discarded.
Myogenic tone was provoked by raising intraluminal pressure from 20 to 120 mmHg in 25-mmHg steps. Each pressure level was maintained for 5 min, and the vessel diameter was then measured. Intraluminal flow (25–125 μl/min, corresponding to shear stresses of 9–45 dyn/cm2 in WT and 4–20 dyn/cm2 in KO) was applied at a pressure of 95 mmHg. Shear stress was calculated by using the formula SS = 4μQ/πr3, where SS is the sheer stress, μ is the viscosity (taken to be 0.007 Poise), Q is the flow (cm3/s), and r is the radius (cm). Flow was maintained for 5 min, and the mean relaxation during this time was measured. Control experiments showed that prolonged application of flow did not give further dilatation. This procedure was repeated with the same vessel after 45 min of incubation with 300 μM Nω-nitro-l-arginine methyl ester (l-NAME). At the end of each experiment, the vessel was relaxed by using Ca2+-free buffer supplemented with 2 mM EGTA. In control experiments, sodium nitroprusside (1 μM) was added to the Ca2+-free buffer. This did not cause additional dilatation, showing that Ca2+-free buffer is sufficient to eliminate tone. Myogenic tone (%) was expressed as [(D1 − D2)/D1]·100, where D1 is the passive diameter in Ca2+-free buffer and D2 is the active diameter.
In separate experiments, SMAs were preconstricted with the selective α1-adrenergic agonist cirazoline (0.03 μM) applied both intraluminally and in the superfusate at 95 mmHg. In these experiments, flow rates were calculated for each vessel to give a shear stress of 25 dyn/cm2, because this resulted in robust dilatation. Flow was applied in the absence of inhibitors, in the presence of l-NAME (300 μM), and in the presence of indomethacin (10 μM). Inhibitors were prepared on the day of the experiment.
Morphometry and immunofluorescence.
Following relaxation in Ca2+-free solution for 30 min, aortic segments and SMAs, adjacent to those used in the mechanical experiments, were fixed with 2% formaldehyde in PBS (pH 7.4) for 30 min. In a subset of experiments, 20 ml fixative was infused through the femoral artery (see In vivo experiments, below) while blood was allowed to leave through the jugular vein. This was done following full arterial dilatation with sodium nitroprusside, which was given as bolus doses (10 nmol in 10 μl) until the blood pressure reduction saturated. Blood pressure was 40 ± 2 mmHg in WT and 34 ± 3 mmHg in KO following nitroprusside saturation (P = 0.08; n = 7 for both). Arteries were then removed and maintained in fixative for another 30 min. Fixed specimens were washed in PBS containing 10% sucrose and methylene blue (2 × 10 min) followed by embedding in Tissue-Tek (Sakura, Zoeterwoude, Netherlands). After freezing, 10 μm cross-sections were obtained in a cryostat. For morphometry, sections were stained with hematoxylin and eosin (Merck, Darmstadt, Germany) or Masson's Trichrome (Biocare Medical, Concord, CA), following the manufacturers instructions, and media thickness, area, and internal circumference were determined by using a computerized image-analysis system (Leica Q500MC). At least four glasses from each mouse were used. Media thickness was obtained from four averaged measurements in positions defined by two perpendicular diameters.
For immunofluorescence, tissue sections were washed (2 × 30 min in PBS), permeabilized with 0.2% Triton-X-100 for 15 min, blocked with 2% bovine serum albumin in PBS for 2 h, and incubated with primary antibodies in the same solution overnight at 4°C. Cav-1 (clone 2297), -2 (clone 65), and -3 (clone 26) antibodies were obtained from BD Biosciences (San Jose, CA) and were used at dilutions of 1:125, 1:200, and 1:1,000, respectively. After being washed in PBS, sections were incubated with Cy5-labeled anti-mouse IgG in PBS with 2% bovine serum albumin for 1 h at room temperature. Nuclei were stained with SYTOX Green (1:3,000; Molecular Probes, Carlsbad, CA). Sections were examined in a Zeiss LSM 510 confocal microscope. Caveolin proteins were detected by monitoring Cy5 fluorescence on excitation at 633 nm. SYTOX Green fluorescence was monitored on excitation at 488 nm.
Transmission electron microscopy.
First-generation SMAs were dissected in Ca2+-free HEPES-buffered Krebs solution, cut into small pieces (∼30, <0.3-mm length), and opened in the longitudinal direction to minimize diffusion distances for the fixative. Following transfer to 1.5-ml Eppendorf tubes, preparations were allowed to settle at the bottom, and the physiological buffer was aspirated and exchanged for fixative containing 2.5% glutaraldehyde in 0.15M sodium cacodylate, pH 7.4. This was repeated once to remove all physiological buffer, and fixation was allowed to proceed overnight. Samples were washed with sodium cacodylate, postfixed for 1 h in 1% osmium tetroxide in sodium cacodylate, and finally washed with sodium cacodylate. They were dehydrated with an ascending ethanol series and were embedded in Agar 100 Resin R1031 medium with acetone as an intermediate solvent. Specimens were sectioned into 50-nm ultrathin sections and were stained with uranyl acetate and lead citrate. Specimens were observed in a Philips CM10 electron microscope operated at 60 kV. Images were recorded on Kodak SO-163 plates without preirradiation at a dose of 2,000 electrons/nm2.
Western blotting using the Cav-3 antibody (see above) was performed as described (27), with dilutions recommended by the manufacturer. SMAs from one mouse gave insufficient protein for reliable quantification. Therefore, SMAs from three mice were combined for each homogenate (n = 1), whereas single aortas and portal veins were used for individual homogenates. Protein concentration was determined with a Bio-Rad protein assay (Bio-Rad, Hercules, CA), and equal amounts of protein were loaded in all lanes. To illustrate equal loading, nontransferred proteins were stained with Coomassie blue, and a section around the actin band is shown.
To assess DNA synthesis independently of hemodynamic conditions in vivo, whole mesenteric arterial trees were subjected to organ culture. SMA trees were dissected under sterile conditions, and blood was removed from the lumen by gentle stroking with a rubber policeman. Following weighing, preparations were cultured in DMEM:Ham's F-12 medium (1:1) with 2% dialyzed fetal calf serum and 10 nM insulin and with the addition of 50 U/ml penicillin and 50 μg/ml streptomycin for 4 days. [3H]thymidine (10 μCi; Amersham Biosciences Europe, Uppsala, Sweden) was added during the last 24 h of culture. On harvesting, preparations were washed in ice-cold PBS and were frozen in liquid nitrogen. Following freeze crushing, preparations were dissolved in 400 μl 0.2 M NaOH and were sonicated twice, and 300 μl was removed. Following two cycles of precipitation with trichloroacetic acid (5% TCA), centrifugation (13,200 g), and washing (5% TCA), the supernatant was discarded and the pellet was resuspended in Soluene. After 2 h, 8 ml of liquid scintillation cocktail was added, and preparations were subjected to scintillation counting (Beckman LS6500, Beckman Instruments, Fullerton, CA). Protein concentration was determined in the remaining homogenate (100 μl) by using the Bio-Rad protein assay. Radioactive counts were divided by the amount of protein and were normalized to WT (100%) in each experiment.
In vivo experiments.
Mice were placed in a small container, and surgical anesthesia was induced with 4% isoflurane (Isoflurane Forene; Abbot Scandinavia, Solna, Sweden) in room air. After mice were transferred to a heating pad, keeping body temperature at 37 ± 1°C with a rectal probe (Temperature Control Unit HB 101/2; Panlab, Barcelona, Spain), anesthesia was temporarily maintained with a small mask. The animals were tracheostomized, and anesthesia was controlled and maintained by mechanical ventilation with air containing 2% isoflurane by using a mouse ventilator (28025; Ugo Basile, Comerio, Italy). Tidal volume was set at 0.35 ml, frequency was set to 98/min, and a positive end-expiratory pressure of 5 mmHg was applied. The right jugular vein was cannulated for infusion purposes. The left femoral artery was cannulated for blood sampling and blood pressure monitoring on a polygraph (Model 7B; Grass Instruments, Quincy, MA).
Mean arterial blood pressure, obtained by low-pass filtering the pressure data with the polygraph, was recorded for at least 45 min before the infusion of l-NAME. l-NAME was given as a bolus (5 μmol in 50 μl), followed by continuous infusion (23 mM at a rate of 7 μl/min), to uphold a constant plasma concentration. l-NAME was infused for 1 h, but pressure stabilized at a higher level within a few minutes. Heart rate was obtained by removing the low-pass filter and increasing chart speed at regular intervals during the experiment. Experiments were carried out in a blinded manner, and genotypes were only revealed following complete analysis.
For plasma-volume determination, mice were allowed to stabilize for 20 min following surgery. Tracer amounts of 125I-HSA (human serum albumin-I-125, RISA, 3 mg/ml, 5 MBq; Amersham Health) were administered in the right jugular vein. To reduce the amount of free iodine, the 125I-HSA was filtered before administration by using Amicon YM-30 (Millipore, Bedford, MA). Immediately after the injection, the catheter was removed and the vein was ligated. To determine the amount of 125I-HSA injected, the vial containing the tracer was counted for radioactivity before injection and was then compared with the amount left in the vial and in the equipment involved. The injected volume was kept at 50 ± 20 μl. Four 10-μl blood samples were collected at 5, 20, 40, and 60 min after injection. The catheter was filled with heparinized saline, which was temporarily removed during sampling. After the last blood sample was obtained, 80 μl blood was drawn for hematocrit measurement. Radioactivity measurements were performed in a gamma counter (Wizard 1480; LKP Wallac, Turku, Finland). The amount of radioactivity in plasma (cpm/ml) was plotted vs. time in a semilogarithmic diagram. The data was fitted by using an exponential regression function (y = a·e−bx), and the tracer concentration was extrapolated to time 0. Plasma volume was finally calculated by dividing the total amount of radioactivity injected into the mouse by the plasma tracer concentration at time 0.
Chemicals and reagents.
Unless otherwise specified, chemicals and reagents were obtained from Sigma (St. Louis, MO).
Mean values ± SE are shown. Student's t-test for unpaired data, or, in the case of multiple comparisons, ANOVA followed by Bonferroni's post hoc test was used to calculate statistical significance.
Caveolin expression in SMA, aorta, and portal vein.
The distribution of Cav-1, -2, and -3 in SMAs was examined by using immunofluorescence. Cav-1 and -2 were readily detected both in the endothelium and in the media of SMAs from WT mice (Fig. 1A). The SMA media, but not the endothelium, stained positive for Cav-3 (Fig. 1A). In keeping with loss of Cav-1, no staining exceeding that with primary antibody omission control was seen in KO [Fig. 1A, top right vs. NP (no primary)]. Consistent with >90% breakdown of Cav-2 (5, 23), staining for this protein was largely absent in KO (Fig. 1A, middle right). Similar to previous findings in KO femoral artery (27), a reduced medial staining for Cav-3 was detected in KO SMAs (Fig. 1A, bottom right). To test whether Cav-3 expression in KO is reduced in other vascular preparations, we compared Cav-3 expression in SMAs, aorta, and portal vein by Western blotting (Fig. 1B). Cav-3 expression was reduced by 50–60% in all cases. Staining of nontransferred proteins on the gels with Coomassie blue is shown below the blots to demonstrate protein loading (Fig. 1B).
Ultrastructure of SMAs.
Electron microscopic examination of SMAs from KO mice did not reveal any conspicuous morphological changes at low magnification, other than an apparent increase in media thickness, with an increased number of cell layers compared with WT (Fig. 2, D vs. A). Focal adhesions (dense bands), extracellular matrix, and overall cell morphology appeared similar at higher magnification (Fig. 2, B and E). Further increases in magnification revealed numerous caveolae in the smooth muscle cell plasma membrane in WT (Fig. 2C). Consistent with almost complete loss of caveolae in aortic and bladder smooth muscle (5, 31), caveolae appeared to be absent in the media of KO SMAs, as illustrated in Fig. 2F.
Increased α1-adrenergic contraction in KO SMAs but not in aorta.
An increased media thickness in KO would be associated with increased arterial force development, which may, however, be masked by increased NO production. To eliminate NO production, arteries were denuded and the NOS inhibitor l-NAME (300 μM) was included. Cumulative concentration-response curves using the selective α1-receptor agonist cirazoline revealed greater force in KO compared with WT at all concentrations exceeding 0.1 μM (Fig. 3A), with no change in potency (EC50 = 91 ± 11 nM in WT and 69 ± 9 nM in KO, P > 0.05). No difference in cirazoline-induced force was detected in the denuded aorta in the presence of l-NAME (Fig. 3B, EC50 = 150 ± 29 nM in WT and 183 ± 25 nM in KO). To address whether increased contraction would also be seen using the physiological ligand on α1-receptors, SMAs were stimulated with noradrenaline (1 μM, ∼80% of Emax). Average force traces from 10 individual preparations are shown in Fig. 3C. The difference between KO and WT was maximal after 7–10 min stimulation and was less pronounced during early and late phases of contraction. Peak force was significantly greater in KO (Fig. 3D). The K+-channel inhibitor tetraethylammonium (20 mM) was then added to optimize conditions for force generation. Noradrenaline-induced force was greater also following K+-channel inhibition (Fig. 3D).
Arterial remodeling in Cav-1 KO mice.
To assess arterial structure quantitatively, vascular sections were fixed and stained with hematoxylin and eosin or Masson's trichrome for morphometry (Fig. 4A). The analysis indicated increased media thickness and media area in SMAs (Fig. 4B). Trichrome staining did not indicate increased connective tissue deposits in KO compared with WT and did not reveal any obvious change in adventitial thickness (Fig. 4A). In the aorta, the internal circumference was increased, whereas the media thickness was unchanged (Fig. 4B). In SMAs, the internal circumference was not significantly different between KO and WT either in immersion-fixed (WT, n = 12; KO, n = 8) or perfusion-fixed (WT, n = 3; KO, n = 5) preparations, and there was no significant difference in circumference between these two fixation methods in either genotype. However, lumen dimensions are highly dependent on pressure and might be affected by fixation conditions. In Fig. 4, the pooled data are shown, but lumen dimensions were also determined in live arteries at defined pressure levels as described below and in Fig. 6.
SMAs were mounted in a pressure myograph, and vessel dimensions were determined by using a CCD camera and an edge-detection system (see methods). Measurements were made in the absence of extracellular Ca2+ to abolish arterial tone. Passive pressure-diameter relationships were determined as shown in Fig. 6A. At an intraluminal pressure of 95 mmHg, lumen diameter was not significantly different in KO (257 ± 9 μm) vs. WT (242 ± 12 μm; n = 11; P > 0.05). From the edge-detection system, wall thickness was estimated as 24 ± 2 μm in KO vs. 14 ± 1 μm in WT (n = 12; P < 0.001), and wall cross-sectional area was estimated as 2.2 ± 0.2 × 104 μm2 in KO vs. 1.2 ± 0.1 × 104 μm2 in WT (n = 12; P < 0.001). Importantly, the wall-to-lumen ratio was greater in KO (0.096 ± 0.006 vs. 0.063 ± 0.006; n = 12; P < 0.001). These measurements refer to the total wall thickness as determined by the CCD camera, so to get an estimate of the media-to-lumen ratio, the media areas, determined histologically (Fig. 4A), were used to calculate the expected media thickness at the pressure levels and diameters shown in Fig. 6A. At all pressure levels above 20 mmHg, the media-to-lumen ratio was higher in KO; e.g., at 95 mmHg it was 0.048 vs. 0.036.
To address whether the changes in arterial structure would translate into an altered L0, circumference-tension relationships in endothelium-denuded SMAs were generated under isometric conditions by using the wire myograph. The relationship between internal circumference and active force on depolarization with high-K+ solution (60 mM) was shifted to the right in KO compared with WT (Fig. 5A), as was the relationship between circumference and passive force (Fig. 5B). Determination of the optimum for force development in each experiment revealed a significant increase of L0 in KO (Fig. 5C). Small arteries in KO mice thus have a hypertrophic muscle layer and exhibit signs of outward remodeling but have increased wall-to-lumen ratio.
Increased rate of thymidine incorporation in SMAs.
To assess DNA synthesis independently of hemodynamic conditions in vivo, whole mesenteric arterial trees were subjected to organ culture, and thymidine incorporation was measured. Consistent with a thicker media, mesenteric artery trees weighed more in KO than in WT (4.2 ± 0.2 vs. 2.7 ± 0.1 mg; P < 0.001; n = 7 and 8) and contained more protein (0.24 ± 0.03 vs. 0.13 ± 0.01 mg; P < 0.01). Thymidine incorporation relative to protein content was increased in KO by 31 ± 14% (P < 0.05; n = 6), as assessed in organ culture.
To test properties of tone in KO compared with WT arteries in an integrated setting, experiments were run in the pressure myograph. In a first series of experiments, pressure was increased in steps from 20 to 120 mmHg to produce myogenic tone. Passive lumen diameter was significantly increased in KO compared with WT over the pressure range 20–50 mmHg (Fig. 6A). Active lumen diameter was significantly increased in KO over the entire pressure range (Fig. 6B). Consistent with earlier findings (1, 7), myogenic tone (%reduction in diameter on activation; see methods) was considerably lower in KO compared with WT SMAs (Fig. 6C). l-NAME increased myogenic tone in KO, but it was still significantly lower than in WT. No effect of l-NAME on myogenic tone was observed in WT.
Flow-induced dilatation and combined tone.
We next examined flow-induced dilatation at 95 mmHg and with the combination of α1-agonist and pressure. Flow caused a rate-dependent reduction of myogenic tone in WT SMAs (Fig. 6D). In KO SMAs that had developed myogenic tone, flow-induced dilatation was absent (Fig. 6D). Treatment with l-NAME conferred on the KO arteries an ability to dilate in response to increased flow but did not affect flow-induced dilatation in WT (Fig. 6E). Moreover, flow-induced dilatation was not different in KO compared with WT following precontraction with the α1-adrenergic agonist cirazoline (combined tone; Fig. 6F). Neither l-NAME nor indomethacin inhibited flow-mediated dilatation under combined tone, suggesting that dilatation was due to a mechanism independent of NOS and cyclooxygenase.
To illustrate differences between tone induced by pressure alone and by the combination of pressure and adrenergic stimulation (combined tone), data at 95 mmHg in Fig. 6, C and F, were replotted in Fig. 7. The difference in myogenic tone between KO and WT was greater than was the difference in combined tone (compare Fig. 7, A vs. B). This was seen both in the absence and presence of flow (not shown). Addition of l-NAME did not eliminate the difference in myogenic tone between WT and KO (Fig. 7A) but eliminated the difference between WT and KO during combined tone (Fig. 7B). The lower myogenic tone was therefore compensated by a relatively greater diameter change in response to adrenergic stimulation, so that the combined tone differed less or not at all.
Plasma volume expansion.
Anesthetized WT (n = 7) and KO (n = 8) mice were injected with tracer amounts of 125I-HSA. The concentration of tracer in plasma decreased in an exponential manner for both WT and KO, with the correlation coefficients for the average curves being 0.96 and 0.95, respectively. The plasma volume was significantly increased in KO (6.0 ± 0.3 ml/100g body wt) compared with WT mice (5.2 ± 0.1 ml/100g body wt; P < 0.05). Hematocrit did not differ between genotypes (WT, 0.402 ± 0.006; KO, 0.405 ± 0.013; P > 0.05).
Blood pressure following NOS inhibition.
Blood pressure and heart rate were recorded in WT and KO mice (Fig. 8A) in the presence and absence of l-NAME. Consistent with previous data (26), blood pressure was similar in WT and KO mice before l-NAME administration (Fig. 8B), whereas heart rate was slightly but significantly increased in KO in the present series (Fig. 8C). The relative increase in blood pressure on administration of l-NAME was 17 ± 3% in WT and 38 ± 6% in KO (n = 11 and 6, respectively; P = 0.001), illustrating the greater influence of NO on blood pressure in KO. Following l-NAME, blood pressure was significantly greater in KO compared with WT, with no difference in heart rate (Fig. 8, B and C). We also observed greater blood pressure variability in KO mice, as evident from the blood pressure records in Fig. 8A. These slow fluctuations disappeared following l-NAME administration.
The high basal NO production (5, 23, 32, 35), the major reduction of myogenic tone (1, 5, 7), and the age-dependent development of cardiac dysfunction (4, 32, 35) in Cav-1 KO mice would be expected to result in lowered systemic blood pressure. This has not been seen to occur in adult KO mice (26). The hypotheses that loss of Cav-1 is associated with arterial remodeling, increased α1-adrenergic contraction, and plasma volume expansion were therefore tested. Consistent with the hypotheses, SMAs were remodeled in KO mice and α1-adrenergic contraction was increased under isometric conditions. Moreover, a modest (14%) increase in plasma volume was detected. Mean arterial blood pressure was normal in KO mice compared with WT, as was also shown in a previous study (26). Following NOS inhibition, blood pressure was higher in KO than in WT, indicating the presence of blood pressure-elevating changes masked by increased NO-mediated dilatation.
Remodeling was present in the splanchnic circulation (present study) and possibly in the hind limb (20, 27). In contrast, the aorta exhibited an increase in circumference but not in media thickness or force development. Although the precise mechanisms of remodeling are debatable, some predictions can be made from the architectural reorganization. As pointed out by Folkow (8), an increased wall-to-lumen ratio is of advantage for controlling vascular diameter and resistance. Indeed, a 30% shortening of the outer muscle layer would increase resistance 7-fold by using the geometry in WT mesenteric arteries (media thickness, 11 μm; radius, 103 μm), whereas resistance would increase 11-fold in KO arteries (media thickness, 17 μm; radius, 103 μm), as calculated by using Poiseuille's law. Thus a more efficient control of vascular resistance conferred on small arteries by increased wall-to-lumen ratio may contribute to a greater blood pressure response to vasoconstrictor stimuli in KO than in WT.
We noted increased wet weight and protein contents in KO compared with WT SMAs and slightly higher rates of thymidine incorporation in organ culture. This is consistent with the negative regulatory role of Cav-1 in cellular growth and proliferation (10–13, 17, 23, 30). This effect may contribute to medial hypertrophy. However, other possible mechanisms may play a role. Arterial dilatation at unaltered blood pressure will lead to increased arterial wall tension, which is a powerful stimulus for growth signaling in smooth muscle (14). Aorta, in contrast to SMA, did not show significant media hypertrophy despite increased diameter in the KO. This may be related to the fact that the aorta is a conduit vessel essentially without resistance function and that elastic fibers rather than smooth muscle cells carry the major part of its wall tension.
Responsiveness to α1-adrenergic agonists in denuded SMAs in the presence of l-NAME was increased in KO arteries. It was previously shown that α1-receptor responses were increased in the denuded femoral artery from KO mice and that this difference was eliminated by inhibition of protein kinase C (27). This suggests that Cav-1 and caveole negatively influence signaling downstream of the α1-receptor. Elimination of this effect in KO arteries would act in synergy with the increased media thickness. As seen in the present study, time courses of α1-adrenergic contraction also differ between WT and KO, and contraction by high-K+ solution was, if changed at all, reduced in KO. In endothelium-intact arteries, Drab et al. (5) found reduced responses to angiotensin II and endothelin-1 in Cav-1 KO but unchanged responses to α1-receptor stimulation. This may be taken to indicate that an increased vasodilating drive from the endothelium is compensated by a specifically increased smooth muscle responsiveness to α1-agonists. Further support for a specific role of caveolae in α1-receptor signaling is the finding that chemical loading of the scaffolding domain peptide from Cav-1 into ferret aorta inhibited contraction induced by phenylephrine (15) but not high-K+ solution. Finally, binding of radioactively labeled α1-agonist to lipid raft fractions containing Cav-1 supports an association of α1-receptors with caveolae in some instances (9, 19). Evidence against a role of caveolae in α1-receptor signaling has, however, also been presented. We (6) and others (22) found that α1-receptor responses resist disruption of caveolae by using cholesterol depletion. One possibility that would accommodate differences between different vascular beds is that caveolae dependence of α1-signaling is dictated by the α1-receptor isoform expressed. There are three α1-adrenergic receptors, α1A, α1B, and α1D, and their contribution to contraction may vary between different arteries. No good pharmacological tools are, however, available to test that hypothesis.
Cav-3 expression was reduced in the absence of Cav-1 in several vascular preparations. This is consistent with data in the femoral artery (27) but differs from results in striated muscle (5). Because caveolae were absent, as seen by electron microscopy, it must be assumed that the contribution of Cav-3 to caveolae formation in the SMA media is minimal. Our ability to detect Cav-3 in venous smooth muscle, in which it is often undetectable, may relate to the use of very sensitive blotting reagents. The possible role played by Cav-3 in smooth muscle may be addressed by using Cav-3-deficient mice.
Myogenic tone stimulated by increased intraluminal pressure (the Bayliss effect) was reduced in SMAs from KO mice, which is in keeping with previous studies (1, 5, 7). Our data reveal that myogenic tone partly recovered after l-NAME. This indicates that increased basal NO release plays a considerable role in the impairment of this hemodynamic autoregulatory mechanism in SMAs. The impairment of myogenic tone remaining after l-NAME may involve reduced RhoA/Rho kinase activation, as suggested by Dubroca et al. (7), and functional activation of Ca2+-activated K+ channels, as proposed by Adebiyi et al. (1). Addition of l-NAME in the presence of α1-agonist at 95 mmHg (combined tone) increased KO tone to the same level as that in identically treated WT arteries. This indicates a greater contribution to tone of α1-adrenergic compared with myogenic mechanisms in KO vs. WT arteries in an integrated setting, such that the entire difference in tone between KO and WT is eliminated in the presence of l-NAME.
The endothelium responds acutely to shear stress by increasing eNOS activity and NO production through a process involving caveolae (25). Accordingly, flow-mediated dilatation in adrenergically stimulated carotid arteries, which in part depends on NO, was recently shown to be impaired in KO mice (33). In SMAs, we found that flow-induced dilatation of myogenic tone is reduced in KO mice. However, flow-induced dilatation recovered when eNOS was inhibited and tone was increased. Furthermore, no difference in flow-mediated dilatation was found when vessels were constricted by activation of α1-adrenergic receptors and pressure. Thus the level of tone is an important determinant of the ability to dilate in response to flow, and this dilatation in the SMA seems to be largely independent of NO production, in accordance with earlier findings of NO-independent endothelium-mediated dilatation of resistance vessels in mice (3).
Plasma volumes and hematocrit values determined in the present study to calculate blood volume in a 25-g mouse yield 2.5 ml in KO vs. 2.2 ml in WT. This represents a 14% greater blood volume in the KO mice. Plasma volume expansion may be secondary to heart failure (16) and is expected to increase ventricular diastolic filling pressure, which indeed has been reported to be increased in KO (32). Importantly, increased ventricular filling would compensate reduced cardiac contractility and would thus improve cardiac output. Activation of the renin-angiotensin system and increased levels of vasopressin may mediate changes in plasma volume and contribute to blood pressure normalization, but to our knowledge concentrations of these mediators in plasma have not been determined in the Cav-1 KO mouse model.
Blood pressure was stated to be reduced in Cav-1 KO mice in one previous study (32), but no data were presented. Moreover, blood pressure was reported to be increased in the lung (35) in Cav-1 KO mice. The present study and our previous work (26) reveal no effect of Cav-1 ablation on mean systemic blood pressure. The effect of l-NAME, however, was doubled in KO vs. WT, indicating a much greater influence of NO on blood pressure. This likely reflects alleviated Cav-1 inhibition of NOS activity (4, 5, 18, 23, 35). Heart rate was slightly increased in the absence of l-NAME but not in its presence, suggesting baroreceptor reflex activation in the former situation.
In conclusion, the present study has shown that arterial remodeling and increased α1-adrenergic contraction in KO mice compensate for a reduced myogenic tone in maintaining blood pressure. Moreover, enhanced NO release caused by loss of Cav-1 impairs myogenic reactivity and plays a greater role for blood pressure as revealed by l-NAME. Finally, KO mice have a greater plasma volume. It may be speculated that remodeling and plasma-volume expansion counterbalance increased NO-mediated dilatation and heart failure, resulting in a largely normal, albeit fluctuating, central blood pressure in the systemic circulation in the absence of Cav-1. However, a wide variety of caveolae-associated mechanisms involving additional cells and tissues may play a role, and blood pressure could be reduced earlier or later in life. Whether changes in arterial geometry and plasma volume are compensatory or reflect a primary regulatory role of Cav-1 is not presently known. Plasma volume expansion occurs in heart failure (16), which develops in KO mice (4, 32, 35). Vascular remodeling, on the other hand, may reflect the role of Cav-1 in regulation of growth and proliferation (10–13, 17, 23, 30).
Supported by the Swedish Research Council [Grants 71X-14955 (K. Swärd), 71X-28 (P. Hellstrand), 04X-08285-17A (B. Rosengren)], by grants from Crafoord (R. Hallmann and K. Swärd), Åke Wiberg (K. Swärd), Magnus Bergvall (K. Swärd), and the Royal Physiographic Societies (R. Hallmann and K. Swärd), by the Heart-Lung Foundation (K. Swärd and P. Hellstrand), by Stiftelsen Konsul Thure Carlssons Minne (K. Swärd), and by a grant from the Torsten and Ragnar Söderberg Foundation to the Vascular Wall Program at Lund University. B. Rosengren was supported by a postdoctoral fellowship from the Norwegian Research Council (ES247409).
We thank Gunnel Roos for genotyping, Ina Nordström for help with the thymidine incorporation study, and Matthias Mörgelin for help with electron microscopy. We also thank Gustav Grände for assistance with plasma-volume determinations.
↵* S. Albinsson and Y. Shakirova contributed equally to this work.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society