Regulation of hydraulic conductivity in response to sustained changes in pressure

Min-ho Kim, Norman R. Harris, John M. Tarbell

Abstract

The present study addresses the effect of a sustained change in pressure on microvascular permeability assessed by hydraulic conductivity (Lp) measurements from microvessels of the rat mesentery. With a microperfusion technique, transvascular filtration (normalized to surface area; Jv/S) and Lp were measured in small arterioles (baseline Lp = 0.26 × 10−7 cm·s−1·cmH2O−1) and venules (baseline Lp = 2.88 × 10−7 cm·s−1·cmH2O−1). The main finding of this study is that step increases in microvascular pressure led to time-dependent alterations of Lp. Immediately after a twofold step increase in pressure, Jv/S increased in proportion to the pressure change. This observation is consistent with Starling's law that predicts filtration proportional to the overall pressure gradient when Lp is constant. However, when Jv/S measurements continued for 60–90 min past the step in pressure, there was an initial decrease in Jv/S for 30 min (“sealing effect”) followed by a substantial increase in Jv/S out to 90 min. The sustained increase in Jv/S suggests an increase in Lp of 36 ± 7% for small arterioles and 42 ± 5% for small venules (P < 0.05 for both). In addition, the increase in Lp in response to an increase in pressure was attenuated significantly by nitric oxide synthase inhibition. These results indicate that a pressure-induced mechanical stimulus (possibly Jv) activates a NO-dependent biochemical response that leads to an increase in hydraulic conductivity.

  • hydrostatic pressure
  • nitric oxide
  • transvascular filtration

according to starling's law, transvascular fluid flux through the vascular wall is regulated either by a change in pressure gradient (hydrostatic or oncotic) or by permeability (oncotic reflection coefficient σ or hydraulic conductivity Lp). It is generally believed that Lp does not change with an acute change in pressure, and this principle has been used in calculations of Lp in micropipette-perfused microvessels, where the slope of the linear regression plot between transvascular filtration normalized to surface area (Jv/S) and pressure is considered to be Lp. This linear relationship has been confirmed in various vessel preparations (8, 16, 17, 21, 23, 29); however, most have been obtained for time periods of ∼1 min or less and relatively little attention has been given to the issue of a change in pressure over an extended time period.

Interestingly, dynamic changes in Lp have been observed in cultured endothelial cells exposed to increases in pressure. Endothelial cells experience a characteristic time-dependent decrease in Lp, the sealing phenomenon, under exposure of continuous elevation of hydrostatic pressure (24, 25, 28). After 30–45 min of sealing, Tarbell et al. (25) observed a significant increase in Lp, mediated by NO, in response to increases in pressure sustained for 5 h.

Previous studies that investigated the effect of pressure on Lp have been performed largely with either in vitro models of endothelial monolayers (25) or isolated large vessels (1, 26). To our knowledge, however, there have been few in vivo studies using intact microvessels to address the consequence of a sustained increase in pressure on microvascular Lp, although it is of great importance in a variety of physiological situations including the fluid shift problem in microgravity (6, 15) and pulmonary edema (19, 27). In addition, recent studies also show the crucial role of NO in regulating microvascular permeability in response to mechanical stimuli such as shear stress (31) or luminal pressure (2, 32). We hypothesize that sustained changes in pressure would alter Lp of microvessels via a NO-dependent mechanism, as was observed in a previous in vitro experimental model (25). To test this hypothesis, changes in Jv were measured at elevated microvascular pressure sustained for 1–2 h in both small arterioles and venules with a microperfusion technique. Nitric oxide synthase (NOS) inhibition experiments were performed to determine whether pressure-induced changes in Lp are mediated by NO.

MATERIALS AND METHODS

Animal preparation.

Animal procedures were approved by the Institutional Animal Care and Use Committees of Pennsylvania State University and the Louisiana State University Health Sciences Center. Male Wistar rats (250–300 g) were housed two or three per cage in a controlled environment (12:12-h light-dark cycle) and were given food and water ad libitum. Rats were anesthetized by an intraperitoneal injection of 135 mg/kg thiobutabarbital (Inactin; T-133, Sigma, St. Louis). The right carotid artery was cannulated for systemic injection of saline during the experiment (∼1–1.5 ml) and euthanasia solution at the end of the experiment. A segment of the small intestine was exteriorized through a midline abdominal incision, and the spontaneously breathing rat was placed on its right side on a Plexiglas board so that a selected section of mesentery could be draped over a glass coverslip glued on a hole centered in the board. The exposed intestine, except for the selected mesenteric section under study, was covered with gauze soaked in bicarbonate-buffered saline (BBS) consisting of (in mM) 131 NaCl, 4.7 KCl, 1.2 MgSO4, 20 NaHCO3, and 3.5 CaCl2. After the board was mounted onto the stage of an inverted microscope, the mesentery and intestine were kept moist with a 2 ml/min superfusion of BBS bubbled with a 95% N2-5% CO2 gas mixture and warmed to 37°C. Rectal temperature was maintained near 37°C by positioning an infrared heat lamp over the rat.

Videomicroscopy.

The mesentery was observed through a ×40 objective (Nikon Plan Apo, 0.95 numerical aperture) with a 100-W halogen light source, and brightfield images were captured with a color camera (ImageStar IS209; Optical Apparatus, Ardmore, PA). The image was directed into a videocassette recorder, and the taped image was used for playback analysis with an image grabber and image processor (Optimas; Media Cybernetics, Silver Spring, MD) for length and diameter measurements with a resolution of ∼0.4 μm.

Microperfusion technique and measurement of Jv.

Unbranched (500–1,000 μm from branch to branch point) third-order arterioles (A3, 25–35 μm in diameter) and small venules (30–45 μm in diameter) with minimal leukocyte rolling and adherence were selected and cannulated in situ with a double- or triple-beveled glass micropipette (inner diameter = 0.84 mm, inner tip diameter = 15–25 μm) filled with BBS solution containing 1% BSA (Sigma) and blue-dyed microspheres (1-μm diameter; Polysciences) that acted as filtration markers during occlusion. The concentration of microspheres in the perfusate was maintained at <0.1 volume %; the 1-μm-diameter microspheres were clearly visible in our videomicroscopy system. The micropipette was connected to a regulated pressure source. To measure Jv/S with a modified Landis technique (14), selected arterioles or venules were occluded with a glass micropipette occluder at a position downstream from the cannulation site. The occluder was positioned over the selected arteriole or venule and carefully lowered onto the vessel by a micromanipulator to compress the lumen. Because the micropipette occluder may damage the vessel at the occlusion site, successive measurements obtained by occluding the same damaged site could overestimate Jv/S. Therefore, each successive occlusion was made ∼5 μm upstream of the previous site, so that any damage incurred by one occlusion would not affect the filtration measurement of the following occlusion.

The filtration rate was calculated by measuring the decreasing volume between the micropipette occluder and a flow marker that was 250–350 μm upstream of the occluder. The occlusion lasted for 20 s, and the Jv/S value was obtained by taking an average of the data between 10 and 20 s. During occlusion, flow markers within the vessel gradually move closer together and toward the occlusion site as the intravascular fluid separating the cells filters across the endothelial barrier into the surrounding tissue. There is also a concomitant change in vessel diameter associated with the myogenic response for an arteriole. We assume uniform circular tube geometry of diameter D and length x, where x is the distance of a marker cell from the occlusion site. The calculation of Jv can then be obtained from a volume balance, where the rate of change in volume (V) between a flow marker and the occluder is equal to the rate at which plasma volume is filtered out, Math(1) However, because both the position of the flow marker and the diameter change, we use the chain rule: Math(2) With V = πD2x/4, the partial derivatives can be computed as ∂V/∂D = πDx/2 and ∂V/∂x = πD2/4. Substituting for these partial derivatives, and then dividing dV/dt by surface area S (=πDx), we obtain Math(3) The second term on the right-hand side of this equation is the form used in experimental protocols (11) where vascular diameter is at steady state (that is, when dD/dt = 0 in the first term on the right-hand side). Jv/S was measured at selected time points from an average of two or three microspheres whose movement was monitored during the vascular occlusion.

Measurement of Lp for cannulated vessels.

The effects of hydrostatic and osmotic pressures on fluid filtration are described by Starling's law, Math(4) where P is the hydrostatic pressure, π is the oncotic pressure, σ is the osmotic reflection coefficient, and the pressure subscripts p and t denote plasma and tissue, respectively. Because the mesentery is exteriorized and superfused with protein-free buffer, tissue hydrostatic and oncotic pressures are negligible. The equation reduces to Math(5) Thus hydraulic conductivity can be expressed as Math(6) The product of σ and π was assumed to be constant during the experiment, in which σ was assumed to be 0.9 and πp of 1% BSA = 1.8 mmHg.

Experimental protocols.

With the technique of micropipette perfusion and then measurement of fluid filtration rate, the effects of sustained changes in transvascular pressure on Lp were assessed. After cannulation, baseline pressure was set at 50 mmHg for arterioles and 20 mmHg for venules, and baseline Jv/S was measured for 30 min without changing perfusion pressure.

To investigate the effect of a pressure change on Jv/S, perfusion pressure was changed from 50 to 100 mmHg for arterioles and from 20 to 40 mmHg for venules. After baseline measurement of Jv/S for 30 min a pressure change was induced, and Jv/S was measured again immediately after the pressure increase. The transvascular filtration rate was measured for a total of 90–150 min after cannulation by observing the movement of microspheres after occlusion. A step increase in perfusion pressure also can result in an increase in shear rate in the microvessels. To investigate the possible role of flow and/or shear in Jv/S measurements, experiments also were performed with regulated low-flow conditions. To regulate flow, the downstream side of the vessel was partially occluded to allow flow at a low but nonzero constant value before and after the step increase in pressure. Flow velocity was measured from the distance that intravascular microspheres moved in a given time. Data were not included if microsphere movement was hindered because of proximity to the vessel wall.

To investigate the effect of NOS inhibition on pressure-induced changes in Lp, venules were superfused with the NOS inhibitor Nω-nitro-l-arginine methyl ester (l-NAME, 100 μM; Sigma). This concentration of l-NAME was used previously to test the effect of NOS inhibition on microvascular permeability in cannulated microvessels (22, 23). After the baseline measurement of Jv/S for 30 min at 20 mmHg of perfusion pressure, pressure was increased to 40 mmHg. Jv/S was then measured for an additional 90 min at higher pressure. At the end of this period, l-NAME was superfused for 30 min, during which Jv/S was measured.

The pressure source set perfusion pressure within the pipette; however, a drop in pressure can occur across the pipette tip during flow. The estimated value of the pressure drop can be obtained with the Hagen-Poiseuille equation: an integral over the length of the pipette tip gives Math where dP is the differential change in pressure across a distance dz along the tip, μ is the dynamic viscosity of the perfusate, and Q0 is the volumetric flow rate into the cannulated vessel. D(z) is the pipette diameter, which is a function of z, tapering from ∼20 μm at the tip (Dtip) to 840 μm at the main body of the pipette (Dpipette) at a length L, ∼1 cm from the tip. The diameter as a function of z can be approximated as Dtip + (DpipetteDtip)(z/L), where z = 0 at the tip. The volumetric flow rate can be expressed in terms of the vessel velocity and cross-sectional area as πvDMath/4, where Dv is the vessel diameter and v is the velocity. Substituting for Q0 and D(z) in the Hagen-Poiseuille equation, and integrating over the length of taper (L) gives a pressure drop Math To measure the perfusion flow rate (Q0) in a few selected experiments, 3-μm-diameter microspheres (Polysciences) were included in the perfusate so that velocity could be measured with an optical Doppler velocimeter (Microcirculation Research Institute, Texas A&M University, College Station, TX). Velocities ranged from 1 to 5 mm/s at a perfusion pressure of 50 mmHg and therefore could have approached 10 mm/s when pressure was increased to 100 mmHg. In Fig. 1A, the estimated drop in pressure across the pipette tip is given as a function of tip diameter and perfusion velocity with the preceding equations. The pressure drop is minimal (1.1 mmHg at 10 mm/s velocity in the perfused vessel) for tips of the size that we used (20 μm). However, the drop would have been substantial if the tip diameter had been <15 μm as shown in Fig. 1A. In the experiments in which we occluded the vessel to restrict flow to ∼0.2 mm/s, and also during complete occlusion (to measure Jv/S), virtually no pressure differential existed across the pipette tip. Figure 1, B, C, and D, give the predicted time course of the pressure steps in the experimental protocols of unrestricted perfusion in arterioles and venules and restricted flow in venules, respectively. The depicted values of pressure include the brief transients during the 20-s occlusions during which the pressure drop across the pipette tip was eliminated.

Fig. 1.

Estimated pressure drop (ΔP) across the tapered micropipette tip based on the Hagen-Poiseuille equation and estimated luminal pressure as a function of time in the experimental protocols. A: estimated pressure drop for various values of micropipette tip diameter and perfusate velocity (v) [parameters used in calculation: inner diameter of micropipette = 0.84 mm, tip length = 1 cm, vessel diameter = 30 μm (used to calculate flow rate through the pipette tip), and perfusate viscosity = 1 cP]. B: time course of luminal pressure for small arterioles in unregulated flow conditions. The calculated change in arteriolar pressure (ΔPa) based on the Hagen-Poiseuille equation is 2.6 mmHg (assuming a flow velocity of 10 mm/s at 100 mmHg with a 15-μm pipette tip) during vascular occlusion. C: time course of luminal pressure for small venules in unregulated flow conditions. The calculated change in venular pressure (ΔPv) based on the Hagen-Poiseuille equation is ∼1.0 mmHg (assuming a flow velocity of 4 mm/s at 40 mmHg with a 15-μm pipette tip) during vascular occlusion. D: time course of luminal pressure for small venules in regulated flow conditions: the continuous partial downstream occlusion allowed flow at a low but nonzero constant value, with an insignificant ΔP across the pipette tip during vascular occlusion.

Statistics.

Unpaired t-tests were used to compare two separate sets of data. Multiple data from individual sets of experiments were analyzed with repeated-measures ANOVA, using Bonferroni corrections (GraphPad Instat, San Diego, CA). Error bars represent SE. Statistical significance was set at P < 0.05.

RESULTS

Experiments to measure transvascular filtration in response to a change in pressure were performed with small venules and arterioles of the rat mesentery. All data were obtained with a microperfusion technique in which baseline perfusion pressure was set to 20 mmHg for small venules and 50 mmHg for small arterioles after cannulation. Thirty minutes after cannulation, measurements of Jv/S gave Lp values of 0.26 ± 0.03 × 10−7 cm·s−1·cmH2O−1 for arterioles (n = 7) and 2.88 ± 0.26 × 10−7 cm·s−1·cmH2O−1 for venules (n = 13).

Effect of sustained increase in pressure on time course of Jv/S and Lp.

Figure 2 shows Jv/S and Lp for small arterioles experiencing a step increase in pressure from 50 to 100 mmHg (or time-control data at a sustained 50 mmHg). All data are given relative to the values obtained at the end of a 30-min baseline period. In time-control experiments, there was a transient decrease in Jv/S during the initial 30 min and then a moderate 14 ± 3% increase (P < 0.05, compared with the 30-min baseline value). In experiments in which pressure was increased by 100%, Jv/S increased immediately by 104 ± 4.5%, followed by a reduction in Jv/S for ∼ 20–30 min and then a sustained increase for the final 30 min (see Fig. 2). The value of Jv/S at 90 min was significantly higher than the value at 31 min (immediately after pressure increase). The immediate 104% increase of Jv/S in response to the pressure step corresponds to the twofold increase in perfusion pressure (from 50 to 100 mmHg) and thus resulted in no significant change in Lp (Fig. 2B). When pressure was increased to 100 mmHg, there was a significant increase in Lp at 90 min (36 ± 7%) compared with the value at 30 min. In time-control experiments, even though there was an increase in Lp at 90 min (14 ± 3%, P < 0.05) compared with the value at 30 min, the value tended to be lower than the initial value at 0 min.

Fig. 2.

Transvascular filtration normalized to surface area (Jv/S) and hydraulic conductivity (Lp) as a function of time for small arterioles. A: Jv/S (%) normalized to the baseline value at 30 min for time-control experiments (n = 3) and experiments with a step increase in pressure from 50 to 100 mmHg (n = 7) at 30 min (Jv/S at 0% = 0.029 ± 0.009 μm/s). B: % change in Lp from the baseline value at 30 min. Baseline Lp at 30 min = 0.26 ± 0.03 × 10−7 cm·s−1·cmH2O−1. Error bars represent ±SE. *P < 0.05.

A similar transient response in Jv/S was observed for venules in response to an increase in pressure from 20 to 40 mmHg as shown in Fig. 3. As in the arterioles, a biphasic response was observed in which Jv/S decreased for 20–30 min after the pressure step and then increased for the remaining 60 min. As expected, the value of Lp at 31 min (immediately after pressure increase) was not significantly different from the value at 30 min (immediately before pressure increase). However, Lp increased by 48 ± 13% at 120 min compared with the baseline value at 30 min, with the final value significantly higher than the values at 31 min and 50 min.

Fig. 3.

Jv/S and Lp as a function of time for small venules (n = 8). A: % change in Jv/S normalized to baseline at 30 min with a step change in pressure from 20 to 40 mmHg (Jv/S at 0% = 0.076 ± 0.01 μm/s). B: % change in Lp normalized to baseline at 30 min. Baseline Lp at 30 min = 2.66 ± 0.34 × 10−7 cm·s−1·cmH2O−1. Error bars represent ±SE. *P < 0.05.

Effect of NOS inhibition on pressure-induced increase in Lp.

To investigate the effect of NOS inhibition on the increase in Lp following a pressure change, l-NAME was superfused onto the mesenteric tissue from which Jv/S and Lp were measured in small venules. Figure 4 shows normalized Jv/S and Lp for venules in which l-NAME was superfused at the end of the experiments. In experiments without l-NAME, the biphasic decrease and then increase in Jv/S concluded with Lp values 42 ± 11% above baseline (Fig. 4B). However, the response was inhibited significantly after 30 min of l-NAME treatment: the final value of Lp after l-NAME was not significantly different from that observed at 60 min (before the Lp increase).

Fig. 4.

Effects of Nω-nitro-l-arginine methyl ester (l-NAME; 100 μM) on Jv/S and Lp for small venules (n = 4). A: % change in Jv/S normalized to baseline at 30 min with a step change in pressure from 20 to 40 mmHg; l-NAME was superfused after 120 min. Jv/S at 0% = 0.082 ± 0.01 μm/s. B: effect of l-NAME on % change in normalized Lp. Baseline Lp at 30 min = 2.86 ± 0.4 × 10−7 cm·s−1·cmH2O−1. Error bars represent ±SE. *P < 0.05.

Pressure-induced changes in Jv/S and Lp under regulated flow.

The results displayed in Figs. 24 were obtained under unregulated flow conditions, dictated solely by the perfusion pressure and vascular resistance. This resulted in different flows before and after the pressure change. To address the effect of flow on Jv/S and Lp measurements during the pressure change, similar experiments were performed under regulated low-flow conditions in small venules in which flow was maintained between 0.1 and 0.3 mm/s by graded downstream restriction. Figure 5 compares time-control experiments of Jv/S in small venules to those experiencing a step increase in pressure from 20 to 40 mmHg. Manipulation of downstream resistance prevented any significant changes in flow due to the pressure increase. The Jv/S (Fig. 5A) and Lp (Fig. 5B) responses during regulated flow were similar to those obtained with unregulated flow (Fig. 3). There was no significant change in Jv/S during the time-control experiments in which perfusion pressure was maintained at 20 mmHg. However, when the pressure was doubled to 40 mmHg at 30 min, Jv/S increased by 97 ± 6%. The increase in Jv/S was followed by the familiar biphasic sealing period and subsequent sustained increase, in which the value at 120 min was significantly higher than the value at 31 min (immediately after pressure increase). At 120 min, there was a significant increase in Lp (P < 0.05, 42 ± 5% increase) compared with the value at 31 min, whereas no significant change in Lp was found in time-control experiments. Baseline Lp for these experiments averaged 4.0 ± 0.7 × 10−7 cm·s−1·cmH2O−1. In addition, the effect of NOS inhibition on Lp in regulated flow was similar to the result observed in unregulated flow conditions. Figure 5B shows normalized Lp before and after superfusion of l-NAME for time-control experiments and experiments with a step increase in pressure from 20 to 40 mmHg at 30 min. With or without the preceding pressure step, l-NAME decreased Lp to a value 15–20% below the baseline value, indicating 1) the baseline influence of NO on Lp, 2) the reversibility of the pressure step-induced increase in Lp, and 3) the role of NO in the pressure step-induced increase in Lp.

Fig. 5.

Jv/S and Lp as a function of time for small venules with regulated low flow. A: Jv/S (%) normalized to the baseline value at 30 min for time-control experiments (Jv/S at 0% = 0.128 ± 0.01 μm/s; n = 4) and experiments with a step increase in pressure from 20 to 40 mmHg (Jv/S at 0% = 0.115 ± 0.02 μm/s; n = 5) at 30 min. B: % change in Lp from the baseline value at 30 min for time-control experiments (n = 4) and experiments with a step increase in pressure from 20 to 40 mmHg (n = 5) at 30 min. Baseline Lp at 30 min = 4.0 ± 0.7 × 10−7 cm·s−1·cmH2O−1. Error bars represent ±SE. *P < 0.05, #P < 0.05 compared with the value at 120 min (20 mmHg); ##P < 0.05 compared with the value at 120 min (40 mmHg).

Figure 6A compares the time course of the Jv/S response (to the pressure step) between experiments with and without regulated flow: no statistical differences were found. However, at each selected time point, when the vessel was completely occluded for the 20 s of the Jv/S measurement, a contrast emerged. In regulated flow conditions, in which flow was restricted to ∼10% of the normal value of ∼2 mm/s, Jv/S was maintained constant for the 20-s occlusion. On the other hand, when flow was occluded in unregulated flow conditions, there was a transient decrease in Jv/S, especially during the initial 15 s (Fig. 6B).

Fig. 6.

Effects of flow on Jv/S measurements for small venules. A: comparison of time course of Jv/S (%) between unregulated flow (n = 8) and regulated low-flow (n = 5) conditions, in which pressure was increased from 20 to 40 mmHg at 30 min and Jv/S was normalized to the 30-min baseline value. B: comparison of the time course of Jv/S measured at 30 min for the 20-s vascular occlusion. Error bars represent ±SE. *P < 0.05.

DISCUSSION

The present study addressed the effect of a sustained change in pressure on microvascular permeability assessed by hydraulic conductivity measurements from microvessels of the rat mesentery. With a microperfusion technique, transvascular filtration was measured from small arterioles and venules. The baseline value of venular Lp reported in this study (2.88 × 10−7 cm·s−1·cmH2O−1) is close to that reported by Rumbaut et al. (mean Lp = 2.3 × 10−7 cm·s−1·cmH2O−1; Ref. 23) with a similar model. The baseline value of arteriolar Lp (0.26 × 10−7 cm·s−1·cmH2O−1) was ∼9% of venular Lp. This result agrees with the previous observations obtained from frog mesentery, in which arteriolar Lp was 6% of venular Lp (29).

The main finding of this study is that sustained step increases in microvascular pressure led to time-dependent increases of Lp and these increases were attenuated by NOS inhibition. Immediately after a step increase in pressure, the increased Jv/S was directly proportional to the pressure change, thus resulting in no significant change in Lp. This result is consistent with previous findings from various microvessel preparations with microperfusion techniques (16, 23, 29) and indicates that the response is purely mechanical, induced only by the increase in hydrostatic pressure gradient and not involving any alteration of permeability coefficients. However, when Jv/S was measured for longer periods of time, there was a transient change of Jv/S even though pressure was maintained at a constant elevated value, indicating an alteration of Lp. Similar transient responses were observed for both small arterioles and venules.

Besides pressure, another possible mechanical factor that could influence our observed response is the luminal flow-induced shear stress that accompanies the pressure increase. The permeability of large solute (31) and fluid filtration rates (12, 29, 30) can be affected by perfusion rate. However, it also should be noted that Neal and Bates (17) did not find a positive correlation between shear stress and Lp. In the current study, the pressure-induced Lp responses under regulated low-flow conditions were not significantly different from those obtained under unregulated flow conditions (Fig. 6A), and this suggests that shear might not be an important factor influencing our Jv/S measurements. However, it should be noted that the free-flow condition before vascular occlusion might influence the transient response of Jv/S during the vascular occlusion (Fig. 6B). As a microvessel is occluded flow decreases from its basal value to essentially zero, and this vessel occlusion includes an abrupt change in flow-induced shear stress that could have a time-dependent effect on Lp (10). Therefore, it is possible that the existence of a transient decrease in Jv/S during occlusion at higher unregulated flow might be due to the larger decrease in shear rate than in lower regulated flow conditions.

A growing body of evidence suggests that NO plays a crucial role in regulating microvascular permeability. Many of these studies have been performed through inhibition of NOS in vivo using individually perfused (8, 21, 23) or autoperfused (7) microvessels or in an in vitro experimental model using cultured endothelial monolayers (3). One interesting finding is that the effect of NOS inhibition on microvascular permeability is dependent on the presence of leukocytes. In individually cannulated and buffer-perfused microvessels, Lp is decreased by NOS inhibition (8, 21, 23). However, when microvessels are perfused with blood-borne elements (23) or autoperfused, the application of NOS inhibitors results in a subsequent increase in Lp; this increase can be reversed with anti-neutrophil serum or an antibody against the CD11/CD18 leukocyte adhesion molecule (7). Our results of NOS inhibition with l-NAME are consistent with previous findings that NOS inhibition can attenuate Lp in buffer-perfused microvessels and also consistent with the previous in vitro finding that the increase in endothelial Lp in response to a sustained increase in pressure is mediated by NO (25).

One interesting observation in the present study is that there was a characteristic time-dependent decrease in Jv/S for the initial 30 min after the pressure step. The mechanism responsible for the sealing effect was revealed recently (5). Mechanical and biological mechanisms were hypothesized to be involved in the phenomenon, and it was suggested that pressure application might seal a small pore pathway for water transport passively during the initial period; on the other hand, biological mechanisms regulate tight junction proteins, e.g., zonula occludens-1, after longer periods of time.

The mechanisms responsible for the subsequent increase in Lp by a sustained change in pressure are not yet clearly understood. One possibility is that endothelium might be ruptured by an increase in pressure after longer periods of time, thereby resulting in high transvascular filtration. However, this seems unlikely because the transient increases in Lp were reversed by the addition of l-NAME, suggesting that the microvascular endothelium was intact during the pressure increase. Another possible mechanism is the activation of a biochemical response induced by the mechanical stimulus. In isolated lungs, an increase in pulmonary vascular pressure significantly increased the capillary filtration coefficient, possibly by activating nonselective stretch-activated cation channels and decreasing endothelial cytoskeletal tone (19, 20). In addition, our observations that increased Lp was attenuated by NOS inhibition also support the hypothesis that a mechanical stimulus might activate a biochemical response. However, there are two important questions that remain unanswered regarding this issue. One is the nature of the mechanical stimulus, and the other is the pathway by which NO is released.

Two possible mechanisms for the pressure-induced response are suggested here. In endothelial cells, a pressure increase can stimulate stretch-activated cation channels and induce Ca2+ influx that induces endothelial NO production (9). A transient increase in Ca2+ also can trigger other signaling pathways that influence microvascular permeability (8, 13). Becker et al. (2) reported that a pressure increase produced NO by increasing either endothelial stretch or luminal surface fluid shear forces, and this played a protective role in ischemic lung injury. Zhang and Pallone (32) also observed increases in cytoplasmic Ca2+ and NO generation in response to increases in luminal pressure in isolated descending vasa recta. It also should be noted that opening of new transcellular pathways or stretch-induced separation of endothelial junctions also contributes to an increase in Lp at much higher pressures that rupture the endothelium (18, 19).

Another possible mechanism to explain our observations is that increased endothelial cleft shear stress driven by increased transvascular flow might mediate NO release. When vascular pressure is increased, an increase in transvascular filtration is driven by a classic Starling mechanism. This increased transvascular filtration is expected to induce increases in shear stresses through the interendothelial cleft. In a previous in vitro study, Tarbell et al. (25) suggested that elevated endothelial cleft shear stress induced by increased transmural flow might be a more important factor in influencing the substantial increase in Lp than a direct effect of pressure or stretch of endothelial cells. This hypothesis was based on the reasoning that the increase in Lp induced by increased endothelial cleft shear stress might be regulated by a downstream signaling pathway similar to the one that controls changes in Lp induced by changes in shear stress on the luminal surface (3). Assuming an endothelial cleft modeled as a slit having parallel walls with laminar flow of a Newtonian fluid, flow-induced shear stress along the cleft surface (τw) can be predicted from τw = (3μ/B)(A/Acleft)(Jv/S) (25), where μ is the viscosity, B is the half-width of the slit, A is the total surface area, and Acleft is the surface area associated with the endothelial cleft. If we further assume that the change in Lp is proportional to the change in Acleft and that μ and B are constant, this simple model applied to the results of Figs. 3 and 4 suggests that the endothelial cleft shear stress at 120 min is almost twofold larger than the baseline value at 30 min before pressure increase. It is well established that an alteration in wall shear stress along the apical surface of vascular endothelium mediates the release of NO (4). Because luminal flow in selected experiments was kept constant at a low value during the pressure increase, it is unlikely that luminal flow-induced shear stress was responsible for NO release during increased pressure. It is plausible, however, that increased endothelial cleft shear stress triggers NO release from endothelium, leading to a substantial increase in Lp.

In conclusion, we have obtained evidence that a sustained increase in pressure can regulate Lp in a NO-dependent mechanism. This mechanism might be similar across the arterio-venular span of the microcirculation and appears to be independent of luminal flow and shear stress.

GRANTS

This work was funded by National Aeronautics and Space Administration Grant NAG3–2746 (J. M. Tarbell and N. R. Harris) and National Heart, Lung, and Blood Institute Grant HL-57093 (J. M. Tarbell).

Acknowledgments

Portions of the data from these experiments were included in a PhD dissertation (M. Kim; Pennsylvania State University, 2004).

Footnotes

  • * N. R. Harris and J. M. Tarbell 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.

REFERENCES

View Abstract