We demonstrated previously that inhibition of endothelial nitric oxide synthase (NOS), using pharmacological inhibitors, attenuated the ionomycin- and ATP-induced increases in microvessel permeability (Am J Physiol Heart Circ Physiol 272: H176–H185, 1997). Recently, the scaffolding domain of caveolin-1 (CAV) has been implicated as a negative regulator of endothelial NOS (eNOS). To examine the role of CAV-eNOS interaction in regulation of permeability in intact microvessels, the effect of internalized CAV on the platelet-activating factor (PAF)-induced permeability increase was investigated in rat mesenteric venular microvessels. Internalization of CAV was achieved by perfusion of individual vessels using a fusion peptide of CAV with Antennapedia homeodomain (AP-CAV) and visualized by fluorescence imaging and electron microscopy. Changes in microvessel permeability were evaluated by measuring hydraulic conductivity (Lp) in individually perfused microvessels. We found that the PAF (10 nM)-induced Lp increase was significantly attenuated from 6.0 ± 0.9 (n = 7) to 2.0 ± 0.3 (n = 5) times control after microvessels were perfused with 10 μM AP-CAV for 2 h. The magnitude of this reduction is comparable with that of the inhibitory effect of Nω-monomethyl-l-arginine on the PAF-induced Lp increase. In contrast, perfusion with 10 μM AP alone for 2 h modified neither basal Lp nor the vessel response to PAF. These results indicate that CAV plays an important role in regulation of microvessel permeability. The inhibitory action of CAV on permeability increase might be attributed to its direct inactivation of eNOS. In addition, this study established a method for studying protein-protein interaction-induced functional changes in intact microvessels and demonstrated AP as an efficient vector for translocation of peptide across the cell membrane in vivo.
- endothelial nitric oxide synthase
- nitric oxide
- individually perfused venular microvessel
- hydraulic conductivity
our laboratory and several other investigators, using pharmacological approaches, demonstrated that the increased nitric oxide (NO) production induced by inflammatory mediators is a necessary signaling pathway to increase microvessel permeability (11, 12, 17–20, 22, 26). Recently, it was reported that caveolin-1, the principal structural protein in the caveolae of endothelial cells, interacts with a wide variety of signaling molecules, and endothelial NO synthase (eNOS) is one of its well-characterized targets (7, 9, 16). It was demonstrated that caveolin-1 inhibits eNOS activity by a direct interaction with eNOS through its “scaffolding domain,” located between amino acids 82 and 101 within endothelial plasmalemmal caveolae (15). A 16-amino acid peptide of Drosophila Antennapedia homeodomain (AP) can facilitate the translocation of caveolin-1 scaffolding domain (CAV) across the cell membrane and can reduce inflammation in vivo (1).
The objectives of our present study are as follows: 1) to apply an AP-linked CAV to individually perfused microvessels to establish a method for the study of endothelial protein-protein interaction-induced functional changes in intact microvessels and 2) to examine the effect of caveolin-1-eNOS interaction on microvessel permeability to further elucidate the role of NO in regulation of microvessel permeability. Delivery of CAV to microvessel endothelial cells was achieved by perfusing individual vessels using a fusion peptide of CAV with AP (AP-CAV). Changes in microvessel permeability were evaluated by measuring hydraulic conductivity (Lp) in rat mesenteric venular microvessels. The effect of internalized CAV on microvessel permeability was investigated before and after exposure of the same microvessel to platelet-activating factor (PAF). The effect of inhibition of NOS by pharmacological agents on PAF-induced permeability increase was also investigated. Such studies provide a parallel comparison with results obtained from application of AP-CAV. Internalization of CAV in endothelial cells forming microvessel walls was illustrated with fluorescence imaging and electron microscopy.
MATERIALS AND METHODS
Animal preparation. Experiments were carried out in venular microvessels in rat mesenteries. All procedures and animal use were approved by the Animal Care and Use Committees at West Virginia University. Female Sprague-Dawley rats (2–3 mo old, 220–250 g; Hilltop Laboratory Animal, Scottsdale, PA) were anesthetized with pentobarbital sodium given subcutaneously. The initial dose was 65 mg/kg body wt, and an additional 3 mg/dose was given as needed. The trachea was intubated, and a midline surgical incision (1.5–2 cm) was made in the abdominal wall. The rat was then transferred to a tray and kept warm on a heating pad. The mesentery was gently taken out from the abdominal cavity and spread over a pillar for measurement of Lp. The upper surface of the mesentery was superfused continuously with mammalian Ringer solution during preparation and experimentation. The temperature of the superfusate was maintained at 37°C and was monitored by a thermometer probe placed at the superfusate dripper and regulated by a digitally controlled water bath. All experiments were carried out in venular microvessels, which were classified as segments where there is convergent flow two to four branches distal to true capillaries. The mean diameter of all vessels used for the experiments was 40 ± 8 (SD) μm (n = 36). Blood flow was brisk in all the vessels selected for experiments, and there was no more than one adherent leukocyte per 100 μm of vessel wall.
Measurement of Lp in single perfused rat mesenteric microvessels. All measurements were based on the modified Landis technique, which measures the volume flux of water across the microvessel wall (3). The assumptions and limitations of the original method and its application in mammalian microvessels have been evaluated in detail elsewhere (3, 14). Briefly, a single venular microvessel was cannulated with a glass micropipette and perfused with albumin-Ringer solution (control) containing 1% (vol/vol) rat red blood cells as markers. A hydrostatic pressure (range 50–80 cmH2O) controlled by a water manometer was applied through the micropipette to the microvessel lumen. A charge-coupled device camera was connected to the microscope, and a video was recorded from a segment of the perfused microvessel (400 μm long) throughout each experiment for data analysis. The initial water flow per unit area of microvessel wall [(Jv/S)0, where Jv is water flow and S is unit area of the microvessel wall] was calculated from the velocity of the marker cell after the vessel was occluded, the vessel radius, and the distance between the marker cell and the occlusion site. Any vessel diameter changes can be precisely measured from the video images, which are incorporated into the Lp calculation. To avoid the effect of vessel compliance on marker cell movement, velocity of the marker cell was calculated after 2 s of occlusion. Microvessel Lp was calculated from the Starling equation as follows: Lp = (Jv/S)/ΔP, where ΔP is the effective hydrostatic and oncotic pressure difference across the microvessel wall. If the tissue hydrostatic and oncotic pressures are assumed to be negligible, ΔP represents the pressure difference between the hydrostatic pressure applied to the microvessel and the effective oncotic pressure generated from albumin in the perfusate (BSA at 10 mg/ml has effective oncotic pressure of 3.6 cmH2O). In each experiment, control Lp was measured first with albumin-Ringer perfusate. Then multiple recannulations were made in the same vessel to measure changes in Lp in response to different testing agents. The mean Lp for each perfusate was calculated from all occlusions during that perfusion period, if Lp is relatively constant in the whole time course. Otherwise, Lp is reported as the means of peak and sustained values, if a transient increase in Lp is observed. To translocate peptide across endothelial membrane, each vessel was perfused with albumin-Ringer solution containing AP or AP-CAV (10 μM) for 2 h. Changes in Lp in response to the testing agent were measured in the presence of the loading peptide.
Fluorescence microscopy. Experiments were carried out with a Nikon 300 Diaphot inverted microscope. Each selected venular microvessel was cannulated and perfused with an albumin-Ringer solution containing 10 μM FITC-labeled AP (F-AP) or AP-CAV (F-AP-CAV) for 2 h in the dark at 37°C. After loading was completed, the vessel was recannulated and perfused with albumin-Ringer solution to wash out F-AP or F-AP-CAV from the vessel lumen. Fluorescence images were acquired by a Hamamatsu digital camera with a ×40 oil objective (NA 1.3) through a Nikon FITC HYQ filter set (excitation 480/40 nm, dichroic mirror 505 nm, and band-pass barrier 535/50 nm) and processed with Universal Imaging software.
Electron microscopy. Biotinylated AP-CAV (B-AP-CAV) was used for electron microscopy to visualize the distribution of internalized AP-CAV in the endothelial cells. Each individually cannulated microvessel was perfused with 10 μM B-AP-CAV in albumin-Ringer solution for 2 h and perfused with albumin-Ringer solution to wash out B-AP-CAV from the vessel lumen. The upper surface of the mesentery was then superfused with a fixative solution of paraformal-dehyde (1%), glutaraldehyde (1.25%), and 5% sucrose in 0.1 M phosphate buffer (pH 7.4) for 20 min. After fixation, a small panel of mesentery containing the perfused microvessel was dissected out and allowed to sit in the same fixative at 4°C overnight. The selected tissue was then rinsed in 0.1 M phosphate buffer and permeabilized with 50% ethanol in H2O. H2O2 (3%) was applied for 15 min to pretreat endogenous peroxidase before immunoperoxidase staining. The tissue was processed with the Vectastain Elite ABC kit (Vector Laboratories, Burlingame, CA) and reacted with diaminobenzidine. The tissues were to remain in ice at all steps. For the negative control, the fixation and immunostaining procedures described above were used, except the vessel was perfused with albumin-Ringer solution without B-AP-CAV. The samples were postfixed in osmium tetroxide (1%), dehydrated in a series of ethanol, transferred into propylene oxide, and embedded in Epon (Ladd LX112). Thin sections were stained with alcoholic uranyl acetate and Reynolds citrate. Micrographs (magnification ×32,400) were taken on a JEOL 1220 transmission electron microscope.
Solutions and reagents. Mammalian Ringer solution was used for dissecting mesenteries, superfusing tissue, and preparing the perfusion solutions. The composition of the mammalian Ringer solution was (in mM) 132 NaCl, 4.6 KCl, 2 CaCl2, 1.2 MgSO4, 5.5 glucose, 5.0 NaHCO3, and 20 HEPES and Na-HEPES. The pH of the Ringer solution was maintained at 7.40–7.45 by adjusting the ratio of Na-HEPES to HEPES. All perfusates used for control and test perfusion contained BSA (10 mg/ml).
1-O-alky1–2-acetyl-sn-glycero-3-phosphocholine (PAF) and Nω-monomethyl-l-arginine (l-NMMA) were purchased from Sigma. AP, the Antennapedia internalization sequence from Drosophila Antennapedia homeodomain (amino acids 43–58, RQIKIWFQNRRMKWKK), was purchased from CN Biosciences (San Diego, CA). AP-CAV, the fusion peptide of CAV (amino acids 82–101, DGIWKASFTTETVTKYWFYR) with AP, was custom synthesized by Tufts University with sequences identical to that published by Bucci et al. (1). Biotin- or FITC-labeled peptides (B-AP-CAV, F-AP, and F-AP-CAV) were synthesized by conjugation of identical biotin or FITC to the NH2 terminus of AP or AP-CAV, respectively. Stock solutions of each peptide (10 mM) were prepared with 100% DMSO. The final concentration of each peptide, 10 μM, was achieved by 1:1,000 dilution of the stock with albumin-Ringer solution. PAF was initially dissolved in 95% ethyl alcohol (5 mM) and further diluted to a final concentration of 10 nM with albumin-Ringer solution. All perfusates were freshly prepared with albumin-Ringer solution before each cannulation.
Data analysis and statistics. Values are means ± SE, except where noted otherwise. Changes in Lp are expressed as the ratio of testing Lp to control Lp. The mean values of Lp (control and test) measured in the same vessel were used as paired data. The significance of the differences within or between groups was evaluated by paired t-test and analysis of variance. P < 0.05 was considered statistically significant.
Effect of PAF on microvessel Lp. The effect of 10 nM PAF on microvessel Lp was examined in seven microvessels. The mean control Lp measured with albumin-Ringer perfusate was 2.7 ± 0.5 × 10–7 c·s–1·cmH2O–1. Lp was transiently increased when each vessel was perfused with albumin-Ringer solution containing 10 nM PAF. The mean peak increase in Lp was 6.0 ± 0.9 times the control value, which occurred during 3–8 min of the PAF exposure. The increased Lp fell to 2.8 ± 0.8 times that of the control after 20 min of exposure to PAF. Results are summarized in Fig. 1.
l-NMMA attenuates PAF-induced Lp increase. To investigate the role of NO in the PAF-induced Lp increase, l-NMMA, an NOS inhibitor, was applied to the vessel lumen before PAF exposure. The mean control Lp of five vessels was 2.2 ± 0.8 × 10–7 cm·s–1·cmH2O–1. After each vessel was perfused with solution containing 100 μM l-NMMA for 10 min, the mean Lp was 2.3 ± 0.7 × 10–7 cm·s–1·cmH2O–1, which was not significantly different from that of the control. Then each vessel was exposed to 10 nM PAF in the presence of l-NMMA. The mean peak increase in Lp was significantly reduced from 6.0 ± 0.9 (PAF alone) to 2.5 ± 0.4 times that of the control (P < 0.01). The reversibility of the effect of l-NMMA on the PAF response was examined in two of the five microvessels. The PAF-induced peak increase in Lp recovered to 6.5 ± 0.2 times the control value after 30 min of perfusion with albumin-Ringer solution to wash out the previous agents from the vessel lumen. Figure 2A shows the time course and magnitude changes in Lp from one of the experiments. Figure 2B shows the comparison of the mean PAF responses with and without l-NMMA.
Effect of internalization of AP on basal Lp and PAF-induced Lp increase. It was reported that a 16-amino acid peptide corresponding to the third helix of AP is internalized by cells in culture (4). This peptide has been used to facilitate translocation of CAV across cell membranes systemically or in aortic rings (1). Before investigating the effect of uptake of AP-CAV on microvessel permeability in individually perfused microvessels, we examined the effects of internalization of AP alone on basal Lp and the microvessel responses to PAF. The mean control Lp measured in five microvessels was 1.8 ± 0.2 × 10–7 cm·s–1·cmH2O–1. After 2 h of perfusion with 10 μM AP, the mean Lp was 1.7 ± 0.4 × 10–7 cm·s–1·cmH2O–1. Then each vessel was exposed to 10 nM PAF in the presence of AP. The mean ratio of the peak Lp increase to that measured before the addition of PAF was 6.2 ± 0.8, which was not significantly different from the mean peak Lp increase measured with PAF alone. These results demonstrated that the uptake of AP modified neither basal Lp nor the vessel responses to PAF. Figure 3 shows the time course of the Lp changes in one of the experiments.
Internalization of CAV to endothelial cells attenuates PAF-induced increase in microvessel Lp. The role of CAV in PAF-induced increases in microvessel permeability was examined in five microvessels. After control Lp was measured, each vessel was perfused with 10 μM AP-CAV for2hto translocate CAV into endothelial cells forming microvessel walls. Perfusion of 10 μM AP-CAV for 2 h did not significantly change the baseline Lp (2.0 ± 0.5 vs. 2.1 ± 0.4 × 10–7 cm·s–1·cmH2O–1) but significantly attenuated the PAF response. The mean peak increase in Lp in response to PAF was reduced to 2.0 ± 0.3 times the control, measured after AP-CAV perfusion. The magnitude of this reduction is comparable with the inhibitory effect of l-NMMA on PAF-induced Lp increase. Figure 4 shows the time course of the Lp changes in a single experiment. Figure 5A shows the comparison of the mean time courses of PAF-induced Lp increases with internalization of AP and AP-CAV. Figure 5B compares the magnitude of PAF-induced peak Lp increases with those measured after application of l-NMMA, AP, or AP-CAV. The PAF-induced Lp increase was not significantly changed after AP perfusion but was significantly attenuated in l-NMMA- or AP-CAV-treated vessels.
We also examined the effect of 2 μM AP-CAV perfusion on the PAF-induced Lp increase. Perfusing vessels with 2 μM AP-CAV for 2 h did not significantly inhibit the PAF-induced Lp increase. The peak increase in Lp was 4.3 ± 1.5 times the control value (n = 3). A significant inhibition of PAF-induced Lp increase (2.4 ± 0.4 times control) was achieved by extending the perfusion time to 4 h (n = 5). To avoid the extended perfusion time in an intact microvessel, a 2-h perfusion with 10 μM AP-CAV was used as the optimum combination of peptide concentration and perfusion period for this study.
Uptake of AP and AP-CAV by endothelial cells forming microvessel walls. Uptake of AP and AP-CAV by endothelial cells in intact microvessels was visualized with fluorescence microscopy using F-AP and F-AP-CAV. Localization of the internalized AP-CAV was illustrated with electron microscopy using B-AP-CAV. Uptake of F-AP and F-AP-CAV was examined in eight microvessels (4 vessels for each peptide). Figure 6 shows representative fluorescence images after vessels were loaded with F-AP or F-AP-CAV. F-AP and F-AP-CAV were taken up by all endothelial cells of the perfused microvessel. In F-AP-loaded vessels, fluorescence was distributed throughout the endothelial cells, which yield a diffuse staining pattern (Fig. 6A). In contrast to the diffuse pattern of F-AP-loaded vessels, punctate distribution of fluorescence was observed in endothelial cells with F-AP-CAV-loaded microvessels, which was clearly illustrated when the focal plane was at the bottom of the endothelial cell layer of the vessel wall (Fig. 6B). These fluorescent spots appear to correlate with the pattern of the caveolar invaginations shown on the endothelial plasma membrane by scanning electron microscopy (21). Figure 7A shows the distribution of B-AP-CAV in endothelial cells of the microvessel wall. The darker spots coating the caveolae indicate that AP-CAV was mainly localized in endothelial caveolae of the microvessel walls. The control micrograph (Fig. 7B) was obtained with albumin-Ringer perfusion but was subjected to the same developing procedures used for B-AP-CAV perfusion. Three vessels of each group were studied for electron microscopy.
Two important findings of this study are as follows. First, we established a new effective approach for the study of protein-protein interaction in vivo using individually perfused intact microvessels. Our results demonstrated that the translocating properties of AP efficiently facilitate the transport of CAV across the endothelial membrane after perfusion of AP-CAV in individually cannulated intact microvessels. More importantly, our results quantitatively demonstrated that internalization of AP alone does not modify basal Lp or the responses of microvessel permeability to inflammatory stimuli. Second, our results showed that internalization of AP-CAV significantly attenuates PAF-induced increases in microvessel permeability. The inhibitory action of AP-CAV is similar to that observed with the use of the NOS inhibitor l-NMMA. These results are consistent with the role of CAV, reported by several studies, as a negative regulator of eNOS in endothelial cells (1, 7, 8, 10, 13). This study provided further evidence to support our previous findings that the inflammatory mediator-induced NO release is a necessary step to increase microvessel permeability (11, 12). The NO-dependent pathway accounts for 70–80% of the PAF-induced Lp increase. The remaining increase may be attributed to other signaling pathways or mechanisms.
AP is an efficient vector for translocation of a biologically active compound across the endothelial cell membrane in individually perfused microvessels. The 16-amino acid polypeptide corresponding to the third helix of AP is able to transport polypeptides and oligonucleotides across biological membranes in cultured cells (5, 24). The internalization process appears to be mediated through direct interaction of AP with membrane phospholipids, which is receptor independent and nonendocytotic (4, 5). Recently, Bucci et al. (1) demonstrated that incubation of mouse aortic rings with 10 μM AP-CAV for 20 h inhibited acetylcholine-induced vasodilation and NO production. In addition, in vivo intraperitoneal delivery of AP-CAV reduced hindpaw and ear inflammation caused by local carrageenan injection and mustard oil application in mice (1).
Our fluorescence images (Fig. 6) demonstrated that, under our experimental conditions, perfusion for 2 h with albumin-Ringer solution containing 10 μM AP-CAV effectively translocates AP-CAV into endothelial cells in the intact microvessel wall, which is sufficient to attenuate the PAF-induced Lp increase. In most of the in vitro studies, the fluorescence-labeled CAV was loaded throughout the endothelial cells without a specific binding location (1). This is the first image that shows the spotted intracellular distribution of CAV in a live vessel using fluorescence microscopy, which is very well correlated with the pattern shown in caveolar invaginations on endothelial plasma membrane by scanning electron microscopy (21). In addition, B-AP-CAV (Fig. 7A) demonstrated the colocalization of caveolin-1 with caveolae in endothelial cells forming the microvessel wall. These results demonstrated that AP is an efficient vector to translocate a biologically active compound across the cell membrane. Compared with the techniques from other in vivo and in vitro studies, application of AP-facilitated internalization of peptide to individually perfused microvessels has some unique features. First, the single-vessel perfusion technique allows the solution containing peptide to have direct contact with endothelial cells lining the intact microvessel walls, which makes uptake of the peptide very efficient. Second, this technique enables repeated cannulation in the same vessel, so that basal Lp can be measured before and after uptake of each control or fusion peptide. Additionally, changes in permeability in response to inflammatory stimuli can be compared with control values measured in the same vessel, which minimizes the discrepancies between individual vessels and animals. Furthermore, microvessel permeability was measured at a known perfusion pressure, and the surface area for water and solute exchange can also be measured precisely. This is important for this particular study, because it enables us to distinguish the changes in microvessel permeability from the volume changes caused by changing flow dynamics and surface area due to eNOS inactivation and NO reduction. Reduced NO production would also promote leukocyte-endothelium interaction in the presence of blood components. Our experiments, conducted in the absence of blood elements, enable the role of internalization of caveolin-1 in the regulation of microvessel permeability to be studied independently from the permeability changes caused by the reduction of NO-induced leukocyte-endothelium interactions.
Internalization of CAV inhibits PAF-induced Lp increase: role of eNOS inactivation. Our previous study demonstrates that NOS inhibitors attenuated the permeability increases induced by ATP, bradykinin, and ionomycin in mesenteric venular microvessels (11, 12). Similar results have been reported for vascular endothelial growth factor-, histamine-, leukotriene C4-, and PAF-induced permeability increases in different types of vessels and animal species (17, 18, 20, 22, 26). Increased NO production has been detected in cultured endothelial cells and in intact vasculature with bradykinin and PAF stimulation (2, 6). Our present study, showing the inhibition of PAF-induced Lp increase by l-NMMA in rat mesenteric venular microvessels, is further evidence supporting the hypothesis that increased NO production is a necessary signaling pathway resulting in increased microvessel permeability. However, because of the nonspecificity of NOS inhibitors used in previous studies, one cannot distinguish the type of NOS that is activated during stimulation and is responsible for the increased NO production.
Several in vitro and in vivo studies demonstrate that eNOS is markedly enriched in caveolae, where it colocalizes with caveolin-1, the primary structural coat protein of caveolae, suggesting a direct interaction between these two proteins (7, 9, 10). Functional studies show that caveolin-1, through binding its scaffolding domain to eNOS, inhibits eNOS activity in a dose-dependent manner (1, 9, 13). In caveolin-1-null animals, eNOS activity is upregulated because of the loss of caveolin-1 inhibition (23). These studies indicated that caveolin-1 is an endogenous inhibitor of eNOS.
As we demonstrated in this study, internalization of AP-CAV significantly attenuated the PAF-induced Lp increase, which indicated that caveolin-1 plays an important role in regulation of microvessel permeability under inflammatory conditions. Exogenous application of CAV in vivo may have the therapeutic potential to prevent the permeability increase during acute inflammation. A recent study by Duran et al. (6) showed an increase in NO production in the superfusate samples after hamster cheek pouch was exposed to PAF. A direct measurement of NO showed a reduction with AP-CAV internalization in cultured endothelial cells (1). In addition, the colocalization of eNOS and CAV has been reported by several investigators (8–10, 13). On the basis of these in vitro and in vivo studies, it is reasonable to speculate that NO is the potential key signaling molecule that contributes to the PAF-induced permeability increases and that the reduction of NO resulting from inhibition of eNOS by the internalized CAV is the mechanism accounting for attenuation of the PAF-induced permeability increase. Because NOS inhibition has shown the inhibitory effect of permeability increases induced by a variety of stimuli (11, 12, 17–20, 22, 26), it is unlikely that the increased NO is unique for PAF exposure or for a specific stimulus. The increased eNOS activity might be a common reaction to inflammatory stimuli, and the activated eNOS-produced excessive NO is a key molecule leading to the increase in microvessel permeability.
Although evidence supports the hypothesis that the CAV-eNOS interaction is the mechanism attributed to the inhibitory action of CAV on the PAF-induced permeability increase, we recognize that CAV can potentially regulate a variety of signal transduction pathways in caveolae. Caveolin was also found in complexes with β1-integrins and may be potentially linked to integrin function (25). We do not exclude the potential role of caveolin-1-integrin interaction in regulation of focal adhesion and microvessel permeability. Further in vivo and in vitro studies are needed to explore this possibility.
Summary. Internalization of a fusion peptide with AP provides a powerful tool for studying cellular protein-protein interaction-induced functional changes in intact microvessels. This study further demonstrated that the stimulated eNOS-derived excessive NO is the common signaling pathway for inflammatory mediator-induced increases in microvessel permeability. Furthermore, application of CAV has a significant inhibitory effect on the permeability increase during acute inflammation.
We thank Dr. Alan R. Light for providing us electron microscopy protocol, Dr. Roger Adamson and Grete Adamson for comments on the electron micrographs, and Dr. Jun Liu for the AP-CAV used for some of the pilot experiments.
Some of the preliminary results were published as an abstract (27).
This study was supported by National Heart, Lung, and Blood Institute Grant HL-56237.
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 © 2004 by the American Physiological Society