Effects of Pressure-Induced Stretch and Convection on Low-Density Lipoprotein and Albumin Uptake in the Rabbit Aortic Wall
The effects of pressure-driven convection and vessel wall stretching in the pressure-related changes in low-density lipoprotein (LDL) and albumin transport across the arterial wall were studied in vitro in freshly excised rabbit thoracic aorta held at in vivo length and pressurized at 70, 120, or 160 mm Hg for 30 minutes. External rigid polyester sleeves of various diameters (4, 5, or 6 mm) were passed around half of the arterial segments in order to prevent vessel distension during pressurization. The intraluminal solution contained 131I-LDL and 125I-albumin. The transmural distribution of relative concentrations of LDL (CLDL) and albumin (Calb) across the wall was determined in wrapped and unwrapped segments using a serial frozen-sectioning technique. In the unwrapped segments, Calb increased uniformly between 70 and 120 mm Hg (P<.0001) but did not change significantly between 120 and 160 mm Hg (0.0063±0.0009 [n=4], 0.0520±0.0055 [n=9], and 0.0620±0.0071 [n=12], respectively). In contrast, CLDL increased markedly both between 70 and 120 mm Hg (P<.001) and between 120 and 160 mm Hg (P<.05) (0.0025±0.0005 [n=4], 0.0234±0.0029 [n=9], and 0.0393±0.0056 [n=12], respectively), with the increase being much more pronounced in the inner than in the outer media. In the segments wrapped with the 4-mm sleeves, both CLDL and Calb did not vary significantly between 70, 120, and 160 mm Hg. In the segments wrapped with the 5-mm sleeves, CLDL increased significantly between 120 and 160 mm Hg, whereas Calb did not vary significantly with increasing pressure. Our results demonstrate that (1) pressure-induced stretching of the arterial wall is a major determinant of arterial mass transport, and (2) pressure-driven convection accentuates LDL accumulation in the inner media, which may explain enhanced atherosclerosis in hypertension.
Experimental, epidemiological, and postmortem studies have clearly demonstrated that both the severity and extent of atherosclerosis are markedly increased in the presence of hypertension.1 2 Changes in macromolecular transport across the arterial wall might be one of the major mechanisms by which hypertension contributes to atherogenesis.3 Several studies have reported an increase in albumin transport across the arterial wall as a result of elevated transmural pressure,4 5 6 7 and we have shown that LDLs tend to accumulate in the inner part of the vessel wall subjected to high transmural pressure.6 However, the mechanisms leading to this focal LDL accumulation have not been elucidated. Acute hypertension results in the distension of the arterial wall, which may account in part for the increase in macromolecular transport induced by elevated transmural pressure.4 6 In addition, increased transmural pressure enhances the convective pathway of mass transport, which could also contribute to the pressure-induced changes in macromolecular uptake by the arterial wall.8 To determine the relative contribution of pressure-driven convection and vessel wall stretching in the pressure-related changes in LDL and albumin transport across the arterial wall, we developed an experimental in vitro model that allows modification of the intraluminal pressure without changing the wall distension in rabbit thoracic aorta, using a rigid external support wrapped around the aortic wall.
Materials and Methods
LDLs (1.025<d<1.050 g/mL) were isolated from fresh human plasma by sequential ultracentrifugation.9 During 1% agarose electrophoresis using the Corning system (CIBA Corning), LDLs migrated as a single band, and no change in electrophoretic mobility was observed compared with the β-lipoprotein band of normal fresh serum.10 Bovine serum albumin (fraction V, Sigma Chemical Co) and LDLs were labeled with 125I and 131I (Amersham France), respectively, by using McFarlane's method modified by Bilheimer et al.11 Free iodine was removed by passing the solution through a Sephadex G-50 column and by extensive dialysis (50 000–molecular weight cutoff at 4°C) against 0.15 mol/L NaCl containing 50 mg/L gentamicin sulfate and 100 mg/L EDTA with four changes over 24 hours. After dialysis, the protein-bound radioactivity precipitated by 10% TCA was 99.7% and 99.5% for 125I and 131I, respectively. The concentration of lipoproteins was determined by the method of Lowry et al.12 Specific activity ranged from 300 to 500 cpm/ng of albumin and from 350 to 650 cpm/ng of protein for LDL.
Male New Zealand White rabbits (2 to 2.5 kg) were anesthetized by intravenous injection of 30 mg/kg sodium pentobarbital via the marginal ear vein. The trachea was intubated, and the animals were mechanically ventilated. The sternum was split lengthwise, and the aorta was exposed between the heart and the diaphragm. The pleural membrane and surrounding fat were then carefully dissected away from the aorta, and the intercostal arteries were cauterized 2 to 3 mm from the aorta. During the whole operation, the surface of the vessel was kept moist by application of saline solution.
The technique used to excise aortic segments has been previously described.6 The aorta was ligated above the diaphragm. A cannula whose width was adjusted to fit the internal diameter of the aorta was inserted above the ligature, pointing toward the heart, tied to the vessel, and connected to a reservoir placed 80 cm above the animal in order to prevent depressurization of the vessel. The reservoir contained Tyrode's solution supplemented with 4% bovine serum albumin and 0.03% Evans blue dye. The animal was killed, and the thoracic aorta was flushed with the Tyrode's solution. A second ligature was placed at the proximal end of the aorta just below the arch, and a second cannula pointing caudally was tied in place at the upstream end of the ligated segment. An additional ligature was then tied around the middle of the isolated segment, and a third cannula pointing caudally was inserted downstream from this ligature.
The caudal segment of the thoracic aorta was excised while held at its in vivo length by clamping the cannulas to an adjustable rig. The intercostal arteries were then ligated close to the aortic wall to avoid leakage. A 15-mm-long polyester wide-meshed sleeve (mesh diameter, 1 to 1.5 mm; ETHICON, Ethnor S.A.) passed around the proximal or distal cannula was then randomly slipped around one half of the arterial segment in order to prevent distension during pressurization of the segment. Sleeves of various diameters were used as described below. The vessel was connected to a pressurization chamber, immersed in Tyrode's solution containing 40 mg/L gentamicin sulfate, maintained at 39°C, and oxygenated with 95% O2/5% CO2. The proximal part of the thoracic aorta was then excised and processed in a similar way. The vessels were then filled at 70 mm Hg with 3 mL Tyrode's solution containing 0.03% Evans blue dye, 4% bovine serum albumin, and both radioactive 131I-LDL and 125I-albumin at a concentration of ≈0.15 and 0.2 mg/mL, respectively. The dye allowed us to test the integrity of the endothelium at the end of the experiment. We have shown in previous studies that this experimental procedure used to excise arterial segments preserved endothelial and medial integrity.6 8
The intraluminal pressure was fixed at 70, 120, or 160 mm Hg for 30 minutes using a pressurization chamber connected to the proximal end of the arterial segment. Three different sleeves were selected to maintain the external diameter of the wrapped arterial segment at a controlled value of 4, 5, or 6 mm. During pressurization, the unwrapped half of the arterial segment was allowed to freely distend as a result of the elevated intraluminal pressure, whereas the distension of the wrapped segment was limited by the sleeve. For each pressure level, the size of the sleeve was selected in order to maintain the arterial diameter below the size observed in the unwrapped segment: 4-mm sleeves at 70 mm Hg, 4- and 5-mm sleeves at 120 mm Hg, and 4-, 5-, and 6-mm sleeves at 160 mm Hg. In preliminary studies, we found that the external diameter of fully relaxed aorta (no longitudinal stretch, no transmural pressure) was 4.08±0.10 mm (n=6).
In order to evaluate the role of the endothelium, an additional series of experiments was conducted under similar conditions using vessels excised as described above. After cannulation, the endothelium was stripped from the luminal surface by gently passing a 4-mm-diameter balloon catheter through the cannula into the lumen of the vessel. We have previously shown that by use of this method, the endothelium was completely removed and that an intact elastic lamina and a normal underlying tissue were maintained with viable smooth muscle cells.13 A 4-mm sleeve was then slipped around one half of the vessel. The vessels were incubated for 30 minutes at 70 or 160 mm Hg, with intraluminal solution containing both radioactive 131I-LDL and 125I-albumin.
The possible effect of wall constriction caused by the sleeve was evaluated in wrapped and unwrapped endothelialized segments incubated for 30 minutes at 70 or 160 mm Hg by transmural distribution of a small tracer, 51Cr EDTA (Amersham France), which is commonly used for the estimation of the extracellular space.14
Estimation of Tracers in the Wall
After incubation, the intraluminal solution was flushed and stored at 4°C. The external diameters of the pressurized wrapped and unwrapped segments were measured using Vernier calipers with an accuracy of 0.1 mm. The vessels were cut from the cannulas, opened axially, rinsed briefly in isotonic saline, and divided into four segments of roughly equal area (two from the wrapped part of the segment and two from the unwrapped part). The presence of Evans blue dye on the luminal surface was checked, and stained parts were discarded. The segments were laid on a slightly greased microscope slide and quickly frozen in a cryostat at −20°C to prevent further diffusion or degradation of the tracers. The edges of the segments were trimmed. The length and width were measured using Vernier calipers, and the surface area of the segments was calculated. En face 20-μm-thick serial sections were cut through the whole thickness of the wall from the intima to the adventitia. The first slice was incomplete and was discarded when it was visually excessively thin. Nonetheless, the thickness of the first section kept for analysis most likely ranged between 10 and 20 μm, inducing a slight underestimation of the tissue radioactive tracer counts in the first section. The boundary between the media and the adventitia was noted by an alteration in the appearance of the section. The volume of each tissue section was calculated from its thickness and surface area.
The sections were placed in precooled test tubes containing 500 μL of 1% albumin solution, and 500 μL of 20% TCA was added to precipitate the protein-bound label. The mixture was centrifuged at 2500g for 15 minutes, and the supernatant was discarded. The same procedure was followed with triplicate 20-μL aliquots of the intraluminal solution.
131I and 125I radioactivities in each test tube were assayed simultaneously for 3 minutes with a double counting procedure on a gamma counter (Kontron GAMMAmatic). Spillovers of 131I into the 125I channel and that of 125I into the 131I channel were corrected by the channel ratio method. The tissue counts for 125I ranged from 100 to 3000 cpm after subtraction of the background, which was ≈20 cpm (maximal value for counting errors was 5.3%, and mean value was 2.7%). Those for 131I ranged from 60 to 1000 cpm after subtraction of the background, which was ≈2 cpm (maximal value for counting errors was 7.3%, and mean value was 3.7%). Percentage of TCA-precipitable 131I and 125I in the intraluminal solution did not change during the course of the experiments.
Relative CLDL and Calb were calculated for each section as the counts per minute per unit volume of wet tissue divided by the counts per minute per unit volume of intraluminal solution. The CLDL and Calb values for the sections cut from a single segment were plotted against their distance from the luminal surface to the adventitia. Because the number of tissue sections varied slightly between arterial segments, the number of 20-μm intervals was chosen as the mean number of tissue sections obtained for a given experimental condition, approximated to the closest number. The relative tissue concentration values were then calculated by linear interpolation when necessary. Average concentration profiles were constructed by averaging the values at equal intervals across the wall, and mean medial CLDL and Calb values were calculated.
51Cr data were analyzed using the same method as described above, except that TCA precipitation was not performed, and results are presented as relative tissue concentrations.
Measurement of Filtration
In order to evaluate the influence of the external rigid support on fluid filtration across the arterial wall, additional experiments were carried out in six rabbits. The thoracic aorta was excised as described above and divided into two segments, one kept unwrapped and one wrapped. Three segments were wrapped with 4-mm sleeves, three segments were wrapped with 5-mm sleeves, and six were studied unwrapped. Each segment was sequentially pressurized at 70, 120, and 160 mm Hg for 60 minutes at each pressure level, and fluid filtration was measured as previously described.15 Briefly, the artery was pressurized using a manometer connected to the artery with a polyethylene extension tube (1.5-mm internal diameter) filled with 4% albumin–supplemented Tyrode's solution. The movement of a meniscus in the tube was measured every 10 minutes. At each pressure level, fluid movement was followed for 60 minutes, during which the pressure was held constant. At the end of each 60-minute pressure level period, the external diameter and length of the pressurized arterial segment were measured using Vernier calipers, and the surface area was estimated assuming cylindrical geometry. The error due to the calculation of the surface area using external diameter instead of internal diameter was estimated to be <10%, assuming a maximal thickness of ≈200 μm with a diameter of 4 mm. The movement of the meniscus was expressed as volume inflow per unit surface area of the artery wall.
A two-factor repeated measures ANOVA was constructed with CLDL and Calb data to test the effects of position within the media (repeated factor) and intraluminal pressure or wrap diameter (crossed factor). When the effect of repeated factor was statistically significant, multiple-factorial ANOVAs were performed on CLDL and Calb data at each position within the media. A one-way ANOVA was conducted on the data of mean medial CLDL and Calb values to test the effects of intraluminal pressure or wrap diameter. When the variations were found to be statistically significant (P<.05), multiple comparisons were performed using Bonferroni's method.16 Comparisons between unwrapped and wrapped segments were done using paired t tests.17 Results are expressed as mean±SEM.
Effects of Transmural Pressure on Vessel Diameter
At 70 mm Hg, the diameter of the unwrapped segments was 5.22±0.08 mm (n=4). This value was raised to 6.39±0.14 mm at 120 mm Hg (n=9, P<.01) and to 6.70±0.18 mm at 160 mm Hg (n=12, P<.01). This latter value was not significantly different from that found at 120 mm Hg.
Transmural Distributions of LDL and Albumin in Unwrapped Segments
Fig 1⇓ shows that in the unwrapped arterial segments, CLDL and Calb values across the arterial wall were significantly increased at 120 and 160 mm Hg compared with those observed at 70 mm Hg. At 120 and 160 mm Hg, Calb profiles were similar and almost uniform across the media (Fig 1b⇓), whereas CLDL profiles showed a steep gradient on the luminal side, with the CLDL in the first sections increasing with transmural pressure (Fig 1a⇓). However, the CLDL relative gradients were similar for the different pressure levels. At 70 mm Hg, CLDL values were lower than the corresponding Calb values at any position across the media. At 120 and 160 mm Hg, CLDL values remained lower than the Calb values in most sections but were either similar or even higher in the first two.
Transmural Distributions of LDL and Albumin in Wrapped Segments
At 70 mm Hg, the transmural profiles of CLDL and Calb in segments wrapped with the 4-mm sleeve were identical to those observed in unwrapped segments (Figs 2a⇓ and 3⇓a). At 120 mm Hg, CLDL profiles were similar in segments wrapped with 4- or 5-mm sleeves, showing a slight luminal gradient, which was smaller, however, than that observed in the unwrapped segments (Fig 2b⇓). Calb profiles at 120 mm Hg were also identical in 4- and 5-mm wrapped segments (Fig 3b⇓). At 160 mm Hg, marked luminal gradients were seen in CLDL profiles of the segments wrapped with the 5- and 6-mm sleeves (Fig 2c⇓), whereas CLDL values in the rest of the media remained unchanged and very low. Calb profiles at 160 mm Hg showed parallel increases with the diameters of the external wraps (Fig 3c⇓), with Calb values decreasing slightly and continuously from the inner to the outer part of the media.
LDL Uptake in the Media
In the unwrapped arterial segments, CLDL values increased significantly with intraluminal pressure (P<.04) and decreased significantly from the luminal to the adventitial side (P<.0001). In addition, a significant interaction between pressure and position was observed (P<.0001); this interaction was due to the more marked effect of intraluminal pressure on CLDL in the inner part of the media than in the rest of the layers. Interestingly, the difference in CLDL between 120 and 160 mm Hg was statistically significant only in the first two inner sections. Mean medial CLDL values increased significantly between 70 and 120 mm Hg (P<.001) and between 120 and 160 mm Hg (P<.05) (Table 1⇓).
In the arterial segments wrapped with the 4-mm-diameter sleeves, repeated measures ANOVA indicated that intraluminal pressure had no significant effect on the transmural distribution of CLDL. However, a significant effect of the position within the media was observed (P<.001): the CLDL values for the first two inner sections were significantly higher than the others. No significant difference was observed between the mean CLDL values at 70, 120, and 160 mm Hg (Table 1⇑).
At 160 mm Hg, wrapping had no effect on CLDL values in the six external sections of the media but had a marked effect on the first two points. CLDL values in these latter sections were significantly lower in the segments wrapped with the 4-mm sleeves compared with those wrapped with 5-mm sleeves, without a significant difference between the arteries wrapped with 5- and 6-mm sleeves. Mean medial CLDL values were significantly lower in segments wrapped with 4-mm sleeves compared with segments wrapped with 5- or 6-mm sleeves (P<.05) (Table 1⇑).
In the segments wrapped with 5-mm sleeves, repeated measures ANOVA showed that CLDL values in the first four sections were significantly increased at 160 mm Hg compared with 120 mm Hg (P<.01). The mean medial CLDL values were also significantly increased at 160 mm Hg (P<.05) (Table 1⇑).
Albumin Uptake in the Media
In the unwrapped segments, Calb values increased with intraluminal pressure (P<.001), but there was no significant interaction between pressure and position within the media. Mean medial Calb values increased significantly between 70 and 120 mm Hg (P<.0001) but not between 120 and 160 mm Hg (Table 2⇓).
In the arterial segments wrapped with 4-mm sleeves, Calb values were not affected by intraluminal pressure and did not vary according to the position within the media. Similarly, mean Calb values were not influenced by intraluminal pressure (Table 2⇑).
At 160 mm Hg, wrapping the aortic segments significantly reduced Calb values in the media (P<.0001) without any significant interaction with the position within the media. Hence, mean medial Calb values were significantly lower in segments wrapped with the 4-mm sleeves compared with segments wrapped with 5- or 6-mm sleeves (P<.01 and P<.001, respectively). Moreover, the difference between the segments wrapped with 5- and 6-mm sleeves was also significant (P<.01) (Table 2⇑).
In the segments wrapped with the 5-mm sleeves, the mean medial Calb values did not vary significantly between 120 and 160 mm Hg (Table 2⇑).
In the unwrapped segments pressurized at 70 and 160 mm Hg, removal of the endothelium enhanced albumin and LDL uptakes (Fig 4⇓). In deendothelialized unwrapped segments pressurized at 70 mm Hg, a CLDL gradient from the luminal surface to the medial-adventitial boundary was observed, whereas the Calb profile showed a peak about the mid media. Mean medial CLDL and Calb values were significantly increased at 160 mm Hg compared with those obtained at 70 mm Hg (Table 3⇓). At 160 mm Hg, the CLDL profile was almost flat in the inner media, but values decreased sharply toward the medial-adventitial boundary. In the segments wrapped with 4-mm-diameter sleeves, both CLDL and Calb profiles were identical at 70 and 160 mm Hg, and mean Calb and CLDL values were not statistically different between 70 and 160 mm Hg (Table 3⇓). Moreover, CLDL values remained very low and were in the range of those observed in intact arteries (Fig 2c⇑). The mean medial CLDL value in deendothelialized segments pressurized at 160 mm Hg and wrapped with 4-mm sleeves was not significantly different from that calculated in intact arteries wrapped with 4-mm sleeves (0.0082±0.0009 [n=5] versus 0.0061±0.0016 [n=4]). Conversely, Calb values were dramatically increased compared with those obtained in intact vessels (Fig 3c⇑). The mean medial value in deendothelialized segments wrapped with 4-mm sleeves was significantly higher than that in intact arteries wrapped with 4-mm sleeves (0.084±0.008 [n=5] versus 0.0079±0.0008 [n=4], P<.001).
51Cr EDTA Experiments
The 30-minute EDTA uptake was determined in wrapped and unwrapped arteries pressurized at 70 and 160 mm Hg in order to verify whether the rigid external wrap caused compaction of the arterial wall. This is unlikely, since EDTA uptake was significantly lower in unwrapped than in wrapped segments. The tissue–relative EDTA concentration values in unwrapped segments were not different at 70 and 160 mm Hg, whereas in 4-mm wrapped segments they increased significantly between 70 and 160 mm Hg (P<.05). In fact, the mean medial EDTA concentration values at 70 mm Hg were 0.11±0.01 and 0.15±0.01 in unwrapped and wrapped segments, respectively (n=6, P<.05). These values at 160 mm Hg became 0.12±0.01 and 0.19±0.02, respectively (n=4, P<.05).
Filtration Flow Rates
In order to evaluate the effect of wrapping the arterial segments on fluid movement across the wall, the filtration flow rates were measured in individual wrapped and unwrapped segments (Table 4⇓). In the unwrapped segments, fluid filtration increased significantly with increasing pressure (P<.01). At a given pressure level, the filtration was significantly decreased in the segments wrapped with the 4-mm sleeve compared with the unwrapped segments (P<.05). Similarly, the filtration was slightly decreased in segments wrapped with the 5-mm sleeve compared with unwrapped segments, but the difference was not statistically significant (Table 4⇓).
The aim of the present study was to determine the relative effects of pressure-induced convection and arterial wall distension on the uptake of LDL and albumin by the arterial wall subjected to high intraluminal pressure. To separate the effect of pressure-induced wall distension from that of pressure-induced convection, we performed experiments in unwrapped arterial segments and in wrapped vessels in which distension was controlled. The main findings reported here are as follows: (1) When the vessel distension is totally impeded using an external rigid wrap (4-mm diameter), neither the uptake of LDL nor that of albumin is augmented significantly, despite an increase in the intraluminal pressure from 70 to 160 mm Hg, suggesting that the effect of intraluminal pressure in unwrapped arteries is mostly due to pressure-induced distension of the vessel wall. (2) At a certain level of distension (such as achieved in the 5-mm wrapped segments), LDL accumulation in the inner media increases between 120 and 160 mm Hg, whereas albumin uptake is not affected by this rise in intraluminal pressure. (3) Maintenance of a low endothelial permeability by the arterial wrap in segments under high pressure partly explains the moderation of albumin transport but is unlikely to account for the unchanged LDL uptake.
Wrapping the vessels altered the stresses acting upon the arterial wall, which have been proposed to be major determinants of the biology of the arterial wall.18 Even though it is clear that changing distension by wrapping the arterial segments was accompanied by changes in wall stresses, the discussion of the present findings will be focused on the effect of distension instead of stresses, because only distension was measured. Inasmuch as the rigid sleeves prevented the distension of the wall when the intraluminal pressure increased, it can be considered that for a given diameter of the sleeve the transmural pressure applied to the arterial tissue remained constant for different levels of intraluminal pressure as a result of increased external pressure applied to the artery. Therefore, the transmural pressure in wrapped segments was different from that in unwrapped segments, where it was nearly equivalent to the luminal pressure. As long as maximal distension was not achieved, transmural pressure and distension were directly correlated, and their respective effects on LDL and albumin transport could not be distinguished. However, when maximal distension was achieved, as in unwrapped vessels pressurized at 120 and 160 mm Hg, the albumin uptake did not change significantly despite the increase in transmural pressure, underlying the importance of distension rather than transmural pressure.
The external wrap used to control the pressure-induced vessel distension might have caused the media to compact, reducing the extracellular space available for diffusion of LDL and albumin. This could account for the reduced macromolecular uptake observed in the wrapped arteries. In order to address this issue, we used labeled EDTA to mark the extracellular space in wrapped and unwrapped arteries.14 Values of mean medial EDTA concentration we obtained were lower than those previously reported by Lever and Jay14 and by us5 using 14C sucrose as marker of the extracellular space. However, in the present experiments, complete equilibrium may not have been achieved, and the label was incorporated only in the luminal solution, whereas it was placed in both luminal and bath solutions in the previous studies. Our results suggest that even at 160 mm Hg, the 4-mm-diameter sleeve did not reduce the extracellular space of the vessel wall. Surprisingly, we observed a significant increase in mean medial EDTA concentrations in 4-mm wrapped segments at both 70 and 160 mm Hg compared with unwrapped segments. This might be accounted for by partial inhibition of 51Cr EDTA efflux from 4-mm wrapped arteries, which was not due to the sleeves themselves, since they have very large meshes (diameter, 1 to 1.5 mm), but to the consolidation of the external layers of the wall, the solid phase of which was subjected to high stresses, whereas the fluid phase was at nearly null pressure.
Effects of Distension on Albumin and LDL Uptakes
In the unwrapped segments, Calb increased significantly with elevated transmural pressure, in agreement with previous studies conducted in vitro4 5 6 7 19 or in vivo.4 20 21 At a constant high intraluminal pressure (160 mm Hg), albumin uptake increased as a function of vessel diameter (Fig 3c⇑). In contrast, when the vessel diameter was kept constant using the 4-mm-diameter external wrap, increasing the intraluminal pressure did not influence the albumin uptake (Table 2⇑). Additionally, in the unwrapped segments, albumin uptake increased significantly between 70 and 120 mm Hg, corresponding to a significant increase in vessel diameter, but between 120 and 160 mm Hg, neither albumin uptake nor vessel diameter was augmented significantly. The relatively flat transmural distribution profile of Calb observed at 120 and 160 mm Hg is consistent with a predominance of convection over diffusion of label across the media.22 23 Therefore, it appears that pressure-induced distension was the primary factor controlling the albumin uptake, which agrees with an earlier report by Duncan et al,19 who studied labeled albumin uptake in isolated canine aortic segments stretched on a holding apparatus and exposed to various levels of stretch and transmural pressure and found that only stretching influenced the albumin uptake by the arterial wall. The mechanisms by which pressure-induced stretching affected the albumin transport involved convection and possibly transendothelial transport, as discussed below.
A striking difference exists between the transmural distribution of albumin and LDL at high pressure in unwrapped vessels, which agrees with our previous findings.6 Indeed, at 160 mm Hg, both albumin and LDL uptake increase, but the albumin transmural distribution is almost uniform across the media, whereas LDL accumulates in the inner media. When arterial diameter was kept constant using an external 4-mm wrap, LDL uptake did not change with the intraluminal pressure. This suggests, as for albumin uptake, a major role of stretching in the increased LDL uptake induced by elevated pressure. However, although the LDL uptake did not vary with increasing pressure from 70 to 160 mm Hg in the 4-mm wrapped arteries, it increased significantly, especially in the inner media between 120 and 160 mm Hg in the 5-mm wrapped arteries, unlike albumin uptake. This result, together with the significant increase in LDL uptake between 120 and 160 mm Hg observed in the unwrapped arteries, which was not the case for albumin uptake, suggests that pressure-induced convection accentuated the effect of distension on LDL uptake.
Prevention of the increase in albumin and LDL uptakes at 120 and 160 mm Hg by the 4-mm sleeve might be partly accounted for by the low fluid filtration obtained under these conditions. The decrease in filtration observed in the wrapped arteries suggests a structural alteration to the extracellular matrix. As a result, the drag coefficient for the solute, which depends upon the matrix geometry, will also have been modified by wrapping the vessel, and we cannot exclude that a reduction in convective transport may also explain our findings.
Role of the Endothelium
The observed lack of effect of intraluminal pressure on the arterial LDL and albumin transport in 4-mm wrapped vessels might be due to the preservation of the endothelial barrier despite elevated intraluminal pressure. Indeed, in minipig aortic segments maintained in a rigid holding apparatus that impeded pressure-induced distension, Fry et al24 showed that endothelial permeability to albumin and LDL is not influenced by transmural pressure per se. Therefore, transendothelial macromolecular transport appears to be not dependent on pressure-induced convection. However, it might be influenced by pressure-induced wall distension, which could account for the difference in LDL and albumin uptakes observed between wrapped and unwrapped segments. Hence, the experiments in deendothelialized arteries were designed to assess the relative contribution of the endothelium and the media to the stretch-related changes in macromolecular transport through the arterial wall observed at elevated intraluminal pressure. The mean medial concentrations of albumin in deendothelialized unwrapped arteries at 70 mm Hg are in agreement with previous findings in vivo; Ramirez et al23 reported that relative tissue albumin concentrations 30 minutes after injection of labeled albumin were 0.06 and 0.10 at 5 and 60 minutes after balloon deendothelialization, respectively. The 10-fold increase in CLDL observed in the present study in deendothelialized arteries compared with intact arteries is similar to previous findings obtained in vivo by Alavi and Moore.25 These authors reported a serum-to-tissue LDL concentration ratio of 0.0037±0.0001 in intact arteries and 0.061±0.009 in deendothelialized arteries. In the 4-mm wrapped deendothelialized segments pressurized at 160 mm Hg, neither LDL nor albumin uptake was significantly different from that obtained at 70 mm Hg. Moreover, in the segments wrapped with 4-mm sleeves pressurized at 160 mm Hg, LDL uptake was not affected by endothelium removal. Conversely, albumin uptake was significantly increased in the deendothelialized wrapped segments compared with intact ones. Mean medial albumin concentrations in 4-mm wrapped segments at 160 mm Hg were increased by >10-fold after endothelium removal but remained significantly lower (2-fold) than the corresponding values in unwrapped deendothelialized vessels. Therefore, preservation of a low endothelial permeability by preventing pressure-induced wall distension was unlikely to be the major cause of the unchanged LDL uptake in wrapped segments at high pressure but might partly explain our findings concerning albumin transport. The unaltered accumulation of LDL after endothelium removal in 4-mm wrapped arteries might be accounted for by the presence of the internal elastic lamina, which acts as a major barrier to large macromolecules.3 26
Albumin and LDL Transport Mechanisms Across the Arterial Wall
The results of the present study show that the effects of acute hypertension on albumin and LDL transport across the arterial wall are controlled by different mechanisms. Albumin uptake by the media depends on both endothelial and medial properties. However, when the endothelium is intact, pressure-induced distension seems to be the major determinant of albumin uptake by the media. This is suggested by our results in wrapped segments, as well as in unwrapped segments at 120 and 160 mm Hg, showing that albumin uptake was insensitive to pressure as long as the vessel diameter was not changed. However, this does not exclude the possibility that the albumin flux through the arterial wall, which was not measured in the present study, did increase between 120 and 160 mm Hg as a result of the increase in fluid filtration, as shown by Lever and Jay27 in the carotid artery wall. Conversely, LDL uptake at elevated intramural pressure appears to be mainly dependent on the medial properties and seems to be influenced by both pressure-induced distension and pressure-driven convection. In addition, our results provide evidence that a substantial amount of LDL can enter the inner media only after a threshold of distension has been achieved. When the vessels were maintained at a small diameter using the 4-mm wrap, LDL uptake remained very low and was not significantly altered by intraluminal pressure. However, more LDLs enter the inner media when the arterial segments are distended by high pressure, as shown in unwrapped arteries pressurized at 120 and 160 mm Hg. These findings are consistent with the hypothesis that LDLs are forced by pressure-driven convection into the media, which in that case behaves as a porous material, whereas albumin can move freely across the media and avoid accumulation in the inner wall because of its smaller size.6 When the media is not stretched, it remains poorly permeable to LDL, whatever the intraluminal pressure applied. But once distension occurs, the media becomes more permeable, allowing the penetration of LDL into the wall as a result of pressure-driven convection and increased endothelial permeability.
We must be careful in extrapolating the present results obtained in vitro to the in vivo situation. The lowest pressure level used in the present work (70 mm Hg) was slightly less than physiological pressure, and diameters of aortas restrained by 4-mm wraps were constricted compared with the physiological diameter. Yet, using a large range of intraluminal pressures and distensions allowed us to reveal the effects of pressure-driven convection and stretching on LDL and albumin uptake by the arterial wall. We must also be careful in extrapolating our in vitro findings obtained under experimental conditions of acute hyperpressure to in vivo chronic hypertension. We previously reported that in an experimental model of chronic hypertension in vivo, elevated arterial pressure per se tended to increase albumin concentration in the rat aortic media, but because of the hypertension-induced structural changes of the media, the net effect was a reduction in transmural concentration of albumin.28
Pressure-induced stretching of the vessel wall is a major determinant of albumin and LDL transport across the arterial wall, and pressure-driven convection specifically enhances LDL accumulation in the inner media, which might explain increased atherosclerosis susceptibility in hypertension.
Selected Abbreviations and Acronyms
|Calb||=||relative concentration of TCA-precipitable albumin|
|CLDL||=||relative concentration of TCA-precipitable LDL|
- Received April 5, 1995.
- Accepted May 20, 1996.
Chobanian AV, Lichtenstein AH, Nilakhe V, Haudenschild CC, Drago R, Nickerson C. Influence of hypertension on aortic atherosclerosis in the Watanabe rabbit. Hypertension. 1989;14:203-209.
Fry DL. Mass transport, atherogenesis and risk. Arteriosclerosis. 1987;7:88-100.
Duncan LE, Cornfield J, Buck K. The effect of blood pressure on the passage of labeled plasma albumin into canine aortic wall. J Clin Invest. 1962;41:1537-1545.
Tedgui A, Lever MJ. Effect of pressure and intimal damage on 131I-albumin and 14C-sucrose spaces in aorta. Am J Physiol. 1987;253:H1530-H1539.
Curmi PA, Juan L, Tedgui A. Effect of transmural pressure on low density lipoprotein and albumin transport and distribution across the intact arterial wall. Circ Res. 1990;66:1692-1702.
Fry DL. Effect of pressure and stirring on in vitro aortic transmural 125I-albumin transport. Am J Physiol. 1983;245:H977-H991.
Tedgui A, Lever MJ. The interaction of convection and diffusion in the transport of 131I-albumin within the media of the rabbit thoracic aorta. Circ Res. 1985;57:856-863.
Havel RJ, Eder H, Bragdon JH. The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest. 1955;34:1345-1353.
Noble RP. Electrophoretic separation of plasma lipoproteins in agarose gel. J Lipid Res. 1968;9:693-700.
Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951;193:265-275.
Bardy N, Karillon GJ, Merval R, Samuel J-L, Tedgui A. Differential effects of pressure and flow on DNA and protein synthesis and on fibronectin expression by arteries in a novel organ culture system. Circ Res. 1995;77:684-694.
Lever MJ, Jay MT. Albumin and Cr-EDTA uptake by systemic arteries, veins, and pulmonary artery of rabbit. Arteriosclerosis. 1990;10:551-558.
Tedgui A, Lever MJ. Filtration through damaged and undamaged rabbit thoracic aorta. Am J Physiol. 1984;247:H784-H791.
Wallenstein S, Zucker CL, Fleiss JL. Some statistical methods useful in circulation research. Circ Res. 1980;47:1-9.
Zar JH. Biostatistical Analysis. Englewood Cliffs, NJ: Prentice-Hall; 1984.
Duncan LEJ, Buck K, Lynch A. The effect of pressure and stretching on the passage of labeled albumin into canine aortic wall. J Atheroscler Res. 1965;5:69-79.
Wu CH, Chi JC, Jerng JS, Lin SJ, Jan KM, Wang DL, Chien S. Transendothelial macromolecular transport in the aorta of spontaneously hypertensive rats. Hypertension. 1990;16:154-161.
Tedgui A, Merval R, Esposito B. Albumin transport characteristics of rat aorta in early phase of hypertension. Circ Res. 1992;71:932-942.
Tedgui A, Chiron B, Curmi P, Juan L. Effect of nicardipine and verapamil on in vitro albumin transport in rabbit thoracic aorta. Arteriosclerosis. 1987;7:80-87.
Ramirez CA, Colton CK, Smith KA, Stemerman MB, Lees RS. Transport of 125I-albumin across normal and deendothelialized rabbit thoracic aorta in vivo. Arteriosclerosis. 1984;4:283-291.
Fry DL, Haupt MW, Pap JM. Effect of endothelial integrity, transmural pressure, and time on the intimal-medial uptake of serum 125I-albumin and 125I-LDL in an in vitro porcine arterial organ-support system. Arterioscler Thromb. 1992;12:1313-1328.
Alavi M, Moore S. Kinetics of low density lipoprotein interactions with rabbit aortic wall following balloon catheter deendothelialization. Arteriosclerosis. 1984;4:395-402.
Penn MS, Saidel GM, Chisolm GM. Vascular injury by endotoxin: changes in macromolecular transport parameters in rat aortas in vivo. Am J Physiol. 1992;262:H1563-H1571.
Belmin J, Michel JB, Curmi PA, Salzmann JL, Juan L, Tedgui A. Reduction of transmural 125I-albumin concentration in rat aortic media by chronic hypertension. Arterioscler Thromb. 1991;11:334-343.