Flow modulates coronary venular permeability by a nitric oxide-related mechanism YUAN YUAN,
HARRIS
J. GRANGER,
DAVID
C. ZAWIEJA,
AND WILLIAM
Microcirculation Research Institute and Department of Medical Physiology, Texas A & M University Health Science Center, College Station, Texas 77843-l Yuan, Yuan, Harris J. Granger, David C. Zawieja, and William M. Chilian. Flow modulates coronary venular permeability by a nitric oxide-related mechanism. Am. J. Physiol. 263 (Heart Circ. Physiol. 32): H641-H646, 1992.-This study demonstrates that flow velocity modulates coronary venular permeability to albumin. Apparent permeability coefficients of albumin (P,) were measured in isolated cannulated coronary venules ranging from 30 to 70 pm in diameter. Hydrostatic and oncotic pressures were controlled while the intraluminal flow velocity was varied. P, at an intraluminal hydrostatic pressure of 12 cmH,O and a flow velocity of 7 mm/s was 4.01 t flow velocity to 10 and 13 mm/s 0.53 x 10m6 cm/s. Increasing augmented the permeability by 33 ,t 14 and 48 t 14%, respectively. The nitric oxide synthase inhibitor, NG-monomethyl+ arginine (L-NMMA) ( 10e5 M), decreased baseline P, and abolished the flow-induced permeability changes. Administration of L-arginine (3 x 10m3 M), a physiological precursor of nitric oxide which reverses the effect of L-NMMA, restored the relationship between flow and permeability. From these results we conclude that 1) flow velocity should be considered as a physical force that potentially modulates permeability of venular exchange vessels in the heart and 2) flow modulates coronary venular permeability via the production of nitric oxide. coronary microcirculation; exchange vessels; ability coefficient; NCI-monomethyl-L-arginine;
apparent permeL-arginine
THE HEART is an organ that depends on a continual
supply of oxygen and nutrients for proper function. Adequate delivery of nutrients to the myocytes depends not only on blood flow but also on the microvascular exchange of nutrients. Conventionally, microvascular exchange is thought to be influenced by physical factors, such as hydrostatic and oncotic pressures, and chemical factors, such as inflammatory mediators (12). Yet, such factors cannot fully explain observations showing that increases in flow are often accompanied by increases in the permeability-surface area product (6, 23). Most studies have focused on a change in surface area during alterations in flow but have not eliminated the possibility that flow may have directly influenced permeability. Because the heart is dependent on fatty acids for metabolic fuel and these substrates are normally bound to albumin, it would seem advantageous for an increase in flow, which typically occurs during enhanced myocardial metabolism, to augment microvascular permeability, i.e., facilitate both delivery and exchange. Therefore, we hypothesized that an increase in flow would augment the permeability of coronary exchange vessels. We tested this hypothesis by studying the relationship between venular permeability and flow velocity in isolated
M. CHILIAN 114
coronary venules. Venules were studied because these vessels are important sites for blood-tissue exchange of water and solutes. Accordingly, variations in venular permeability have the potential to alter exchange of water, protein, and nutrients. Because of the well-established link between flow and the production of nitric oxide (NO) (5, 27), and our finding that permeability and flow are also associated, we further speculated that NO is involved in flow-induced permeability changes. MATERIALS General
AND METHODS
Preparation
Pigs weighing 9-13 kg were sedated with ketamine (2.5 mg/kg im) and Rompun (2.25 mg/kg im), anesthetized with Nembutal (25 mg/kg iv), and heparinized (250 units/kg iv). Following a tracheotomy and intubation, the animals were ventilated. A left thoractomy was performed, and the heart was electrically fibrillated, excised, and placed in 4°C physiological saline. The coronary sinus was cannulated, and 3 ml india ink-gelatin-physiological salt solution were infused to clearly define venular microvessels. This solution was prepared by adding 0.2 ml of nondialyzed india ink (Koh-I-Noor, Bloomsbury, NJ) and 0.35 g of porcine skin gelatin to IO ml of warm physiological salt solution and filtering through P8 filter paper (Fisher Scientific, Pittsburgh, PA). This perfusate does not affect microvascular endothelial reactivity (19). Solutions
and Perfusates
The albumin-physiological salt solution (APSS) used for venule dissection had the following composition (in mM): 145.0 NaCl, 4.7 KCl, 2.0 CaCl,, 1.17 MgSO,, 1.2 NaH,PO,, 5.0 glucose, 2.0 pyruvate, 0.02 EDTA, and 3.0 3-(N-morpholino)propanesulfonic acid (MOPS) buffer. After adding 1% bovine serum albumin, the solution was buffered to a pH of 7.40 at 4°C and then filtered through a Millex-PF O.&pm filter unit (Millipore, Bedford, MA). The APSS used to perfuse the vessels had the same composition as mentioned above but was buffered to a pH of 7.40 at 37°C and filtered. The test solute for permeability measurements was fluorescein isothiocyanate-bovine albumin (FITC-albumin) (Sigma, St. Louis, MO). The test solutions were prepared by adding 5 mg FITC-albumin to 100 ml APSS. Isolated
Venule
Preparation
The first step of the dissection procedure was visualization of a suitable venule (length, 0.8-1.2 mm; diameter, 30-70 pm) with a Nikon dissecting microscope. The venule and a small amount of myocardium were excised from the surrounding tissue and then moved to a chamber containing APSS (4°C) for further dissection. Dumont forceps and iridectomy scissors were used to remove the tissue surrounding the venule until the desired
0363-6135/92 $2.00 Copyright 0 1992 the American Physiological Society Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
H641
H642
FLOW
MODI
‘LATES
CORONARY
length of venule was cleared of cardiac tissue. The exposed portion of the vessel was then cut and transferred to the cannulating chamber, which was mounted on a Zeiss Axiovert microscope and gradually warmed to 37°C. The isolated venule was cannulated on both the inflow and outflow ends with two micropipettes (inflow pipette tip: OD 40 pm, ID 30 pm; outflow pipette tip: OD 40 pm, ID 20 pm) and secured with 11-O suture (Alcon, Fort Worth, TX). A third pipette (tip: OD 20pm, ID 16 pm) was inserted into the inflow pipette. The vessels were perfused with either APSS through the outer inflow pipette or APSS containing fluorescently labeled test solutes through the inner inflow pipette (Fig. 1). The cannulating micropipettes were connected to three reservoirs so that intraluminal pressure and flow could be adjusted independently by simultaneously changing the heights of inflow and outflow reservoirs in equal increments. For instance, setting the inflow reservoir at 20 cmH,O and the outflow reservoir at 0 cmH,O produced an intraluminal pressure of 12 cmH,O and an intraluminal flow velocity of 7 mm/s. (Intraluminal pressure was slightly higher than the average of inflow and outflow pressures because of the difference of resistance between the tips of the inflow and outflow pipettes.) Simultaneously raising the inflow reservoirs to 30 cmH,O and decreasing outflow reservoir to -10 cmH,O increased intraluminal flow velocity to 13 mm/s while maintaining a relatively constant intraluminal pressure (intraluminal pressures increased slightly to 15 cmH,O). Because the apparent permeability is influenced by pressure, possible effects of these small changes in pressure during increases in flow velocity were corrected by factoring out the pressuredependent flux of protein. To accomplish this, the permeability to albumin was measured at constant flow during elevation of inflow reservoirs 20 cmH20 APSS FlTC tflow reservoir 0 cmH20
VENULAR
pressure to 12 and 15 cmH,O, which was produced by elevating both the inflow and outflow reservoirs. These values were then subtracted from the apparent permeability observed at the two high flow rates. The image of the vessel (x20) was projected onto a siliconintensified target camera and displayed on a high-resolution monochromatic video monitor. The vessel diameter was measured on the monitor with a calibrated video caliper. The intraluminal pressure was measured through a pressure transducer connected to the inner inflow pipette when the pipette tip was advanced into the venular lumen. The intraluminal flow velocity was measured with an optical Doppler velocimeter (Microcirculation Research Institute, Texas A & M University) (4) by perfusing the vessel with a 1% red blood cell suspension. The preparation was discarded if leakage of the fluorochromes occurred. Measurement
II
l-----JL,
1
‘i-r’~e
7
inner pipette
outer pipette
outflow pipette
Fig. 1. Schematic diagram showing pipettes and reservoirs that control intraluminal flow and pressure of cannulated venule. Venule was cannulated with 2 inflow pipettes (outer and inner) and an outflow pipette. The 2 inflow reservoirs connected to inner and outer pipettes were mounted toge ther so that they could be raised or lowered simultaneously. Increasing inflow reservoirs and decreasing outflow reservoir in equal increments increased intraluminal flow velocity while intraluminal pressure was maintained at a relative constant level . Simultaneously increasing inflow and outflow reservoirs in equal increments increased intraluminal pressure wi thout significant changes in flow velocity. This system enables precise measurements of venular permeability under conditions in which forces that govern permeability and factors that influence permeability are controlled. APSS, albumin-physiological salt solution; FITC, fluorescein isothiocyanate.
of Venular
Permeability
The permeability of the vessel was measured by the ratio of the transmural flux of fluorescently labeled molecules per unit surface area per unit time to the transmural difference of solute concentration, as developed by Huxley and Curry (8). With the use of an optical window of a video photometer (Microcirculation Research Institute, Texas A & M University) positioned over the venules and adjacent space on the monitor, the fluorescent intensity from the window Twas measured. In each measurement, the isolated venule was first perfused with APSS through the outer inflow pipette to establish a baseline intensity. The venular lumen was then rapidly filled with fluorochromes by switching the perfusion to the inner inflow pipette, which produced a step increase and subsequently a gradual increase in the fluorescent intensity. There was a step decrease of intensity when the fluorochromes were washed out by switching the perfusion to the outer inflow pipette. The permeability coefficient to albumin (P,) was calculated using the equation P = (l/AI,)(dI,ldt),(r/2), where AI, is the step increase in fluorescent intensity, (dI,ldt), is the initial rate of increase in intensity as solutes diffuse out of the vessel and into the extravascular space, and r is the venular radius. Measurements were repeated at l- to 2-min intervals. Experimental
APSS
PERMEABILITY
Protocols
Effect of flow on venular permeability. The baseline P, was measured in 12 isolated coronary venules (12 vessels from 12 pigs) under an intraluminal pressure of 12 cmH,O and flow velocity of 7 mm/s. The relationship between P, and flow velocity was tested by measuring P, at a series of intraluminal velocities. To produce a series of flow velocities while maintaining the intraluminal pressure relatively constant, the heights of inflow and outflow reservoirs were adjusted as follows: 1) inflow 20 cmH,O, outflow 0 cmH20; 2) inflow 25 cmH,O, outflow -5 cmH,O; and 3) inflow 30 cmH,O, outflow -10 cmH,O, producing flow velocities of 7, IO, and 13 mm/s, respectively. These velocities are in the range of in vivo measurements of flow velocities in small coronary venules (I). In four preparations, a 1% red blood cell suspension was introduced into the lumen, and flow velocity at various pressures was measured with an optical Doppler velocimeter. Effect of NO on flow-related venularpermeability. The role of NO in flow-dependent permeability changes was studied by comparing the effect of flow on venular permeability during control conditions (n = 12) and after administration of NO synthase inhibitor NG-monomethyl+arginine (L-NMMA; low5 M) in the suffusion bath for 20-30 min in 8 of the above 12 vessels. To establish that the effects of L-NMMA were specific, P, was then measured in the presence of L-arginine (3 X 10e3 M; n = 3), which is reported to reverse the effects of the antagonist
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
FLOW
MODULATES
CORONARY
VENULAR
(26). To eliminate the possibility that the effects of L-NMMA on permeability were mediated by its charge, the same concentration of NG-monomethyl-D-arginine (D-NMMA), which is a biologically inactive isomer of L-NMMA, was applied for comparison. Data
H643
PERMEABILITY Y
Control
v
L-NMMA
Y
L-arginine
Analysis
P, was measured two or three times at each flow velocity and averaged. The mean P, at each flow level was normalized to the control P, at 7 mm/s, i.e., (P, at certain flow/P, at 7 mm/s) X 100, to obtain the relationship between intraluminal flow velocity and percentage changes of P,. Regression analysis was applied to examine the statistical significance. The differences were considered significant when P < 0.05. All the data were presented as means t SE. RESULTS Effect of Flow on Venular
Permeability
Under control conditions, the P, obtained from 12 vessels was 4.01 2 0.53 X lOA6 cm/s at an intraluminal pressure of 12 cmH20 and a flow velocity of 7 mm/s. Increases in flow velocity to 10 and 13 mm/s augmented the uncorrected permeability by 48 t 14 and 71 t 14%, respectively (Fig. 2A, dashed line). Intraluminal pressure also increased modestly during the increase in flow (2 and 3 cmHzO, respectively, from the control pressure of 200
50 I
0 5
, 7
Flow
I
1 9
Velocity
I
1
I
11
I
’
13
1
15
(mmkec)
Fig. 3. Albumin permeability in isolated coronary venules increased 48% as intraluminal flow increased 6 mm/s. Effect of flow on permeability was abolished by treating same venules with NG-monomethylL-arginine (L-NMMA; lo+ M). L-Arginine (10v3 M) in the presence of L-NMMA restored the flow-induced permeability change. All data have been corrected for effects of pressure (refer to Fig. 2 for details).
A cu n
180
E 6
140
z b0 F
120 160
2
80
! 6
I 8
Flow
I 10
Velocity
I 12
1 14
(mmkec)
12 cmH20). To eliminate the potential influence of these small intraluminal pressure changes on P,, we measured P, at the three pressures during constant flow to formulate a relationship between P, and pressure (Fig. 2B). The effects of pressure were then subtracted to provide the pressure-independent relationships between flow velocity and P, (Fig. 2A, solid line). Importantly, the corrected P, still increased significantly as a function of flow velocity: P, increased 33 and 48% as intraluminal velocity increased 3 and 6 mm/s, respectively, indicating that venular permeability is flow dependent. Effect of NO on Flow-Related
Q e
180
0 L
160
I
80
i 11
I 12
Intraluminal
I 13
I 14
Pressure
'
I
15
(cmH
m
I
16
*O)
Fig. 2. A: apparent permeability coefficient of albumin (Pa) increased as a function of flow velocity in isolated coronary venules. Relationships between P, and flow velocity before (dashed line) and after correction (solid line) are shown. Correction is to factor out effects of slight increases in intraluminal pressure that occurred during increases in flow velocity (intraluminal pressures measured at flow velocities of 7, 10, and 13 mm/s were 12, 14, and 15 cmH,O, respectively). B: relationship between P, and pressure ranging from 12 to 15 cmH20, which was obtained at constant flow by elevating both inflow and outflow pressures by identical amounts. This slope serves as correction factor to subtract effect of small pressure changes in curve of P, vs. flow shown in A.
Venular
Permeability
After L-NMMA treatment (10u5 M; n = 8), the baseline P, decreased from the control value of 4.01 t 0.53 to 3.15 t 0.58 x lo+ cm/s, and the effect of flow on permeability was abolished (Fig. 3). Two-way analysis of variance showed a significant difference between the two curves (P < 0.05). L-Arginine (3 x 10e3 M; n = 3) not only restored the relation between flow and permeability but increased P, 2-3 times over control. D-NMMA (low5 M) did not alter the resting permeability or the flow-induced increase in permeability. Under control conditions, L-arginine alone did not influence permeability. DISCUSSION
This study demonstrates that flow, via the production of NO, is capable of modulating permeability of coronary venules. The apparent permeability coefficient of albumin was measured in isolated coronary venules ranging from 30 to 70 ,um in diameter, in which hydrostatic and oncotic pressure were controlled, but intraluminal flow velocity was varied. P, increased as a function of
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
H644
FLOW
MODULATES
CORONARY
VENULAR
PERMEABILITY
greater than those obtained in isolated frog vessels (9) partially because of the differences between temperatures in the two studies (37 vs. 15°C). On the basis of these arguments, it seems unlikely that the vessels were functionally damaged during isolation and perfusion procedures. It is our belief that the relatively high P, was not an artifact related to damage, but rather reflects the highly permeable nature of the coronary exchange vessels. As stated above, the relatively high permeability of Methodological Considerations coronary venules most likely reflects a physiological charThe apparent permeability coefficient of albumin was acteristic of coronary exchange vessels. Laine and Granger (20) showed that the transvascular flux of fluid measured from the ratio of transvenular flux of fluorescent-labeled solutes per unit surface area per unit time to and protein in the myocardium was at least one order of the transmural difference of solute concentration. This magnitude higher than in skin or skeletal muscles. This analysis assumes that albumin crosses the venular wall high efflux was not explained fully by the larger exchange exclusively by diffusion. Because macromolecular ex- surface area in myocardium because the lymph-to-plasma travasation is also coupled to water filtration, the appar- ratios for total plasma protein (0.8) and for albumin (0.9) ent permeability coefficient of albumin obtained in this were also higher than those in the other tissues such as study is an overestimation of the true permeability of skeletal muscle (29). Pilati (24) found that coronary vascoronary venules. Because we took precautions to stabi- cular reflection coefficient for albumin was relatively low lize the transmural pressure and account for pressure- (0.59 t 0.05) when compared with that of skeletal muscle dependent contributions, the changes in apparent P, re- and mesentery (7). An explanation for the comparatively high coronary permeability is that the heart has higher flect alterations in albumin permeability. We would like to point out that our measurements of metabolic demands than other tissues, and the relatively permeability reflected greater transmural flux and not high metabolism necessitates a rather rapid exchange of nutrients and substrates in the coronary exchange altered vascular filling. This is important to highlight because in the formula to calculate permeability, a change vessels. The flow velocities employed in this study were 7-13 in AI, (step change in fluorescence associated with filling) and/or change in (dIJdt). (albumin flux) alters the cal- mm/s, which are within the normal physiological range of culated permeability. Filling of the vessels (AI,) was not coronary venular flow velocity (1). This range of flow changed; rather, only the flux values (dIJdt), were velocity applied in the vessels ranging from 30 to 70 pm in changed by the interventions. Moreover, venular diame- diameter produced a shear stress in a range of 3.85-16.68 ter did not change during the experimental maneuvers. dyn/cm2. Shear stress (y) was calculated by y = ps, where Thus we can state with conviction that we have studied p is viscosity of the testing solution, which was measured as 0.77 CP in our study, and s is wall shear rate calculated changes in venular permeability as opposed to artifacts associated with altered filling or diameter changes. bY s = 5 x centerline velocity/vessel diameter (2). One P, measured at intraluminal pressure of 12 cmHZO and concern is that alteration of intraluminal velocity, and flow velocity of 7 mm/s was 4.01 t 0.53 X lo+ cm/s, thus shear stress, may have induced changes in vessel which is one order of magnitude larger than that in frog diameter, which would compromise permeability meamesenteric capillaries (4.7 t 1.1 X 10e7 cm/s at 8.1 t 3.5 surements. We did not find any alterations in vascular cmH20) (9). We would argue that this high baseline per- caliber during variations in intraluminal flow velocity. meability was not caused by vessel damage during isola- Importantly, the coronary venules we studied consisted tion because of several observations and facts. 1) Mea- primarily of a single layer of endothelial cells and several surement of P, in isolated hamster cheek pouch venules pericytes; thus they most likely had little active tone. using similar techniques was 2.1 x 10e6 cm/s, which was Therefore, changes in permeability were not due to comparable to that obtained from in situ preparations changes in venular diameter, but were due to the effects of (1.3 x 10m6 cmHZO) (8). 2) P, was stable throughout the flow velocity. experiments and was reversibly affected by the interventions, suggesting that the barrier function of the vessel Flow Modulates Coronary Venular Permeability wall was not severely damaged, and was not progressively A recent study indicated that albumin permeability of deteriorating. 3) Administration of indomethacin, an bovine aortic endothelial monolayer is shear dependent anti-inflammatory agent, did not alter the P, of the ves- and reversible (16). It is difficult to extrapolate results sels, which implies that the experimental procedures did from cultured aortic endothelium to factors that modunot cause prostanoid production associated with damage. late exchange processes in the microvasculature. Our 4) P, was measured at a normal body temperature (36- study suggests a role of flow velocity in regulation of 37°C). Our preliminary experiments showed that in- venular permeability. Shear stress influences physical and biochemical charcreases in temperature from 15 to 36-37°C increased P, two- to threefold in isolated porcine coronary venules, acteristics of endothelial cells (10, 11, 16) and, more imand this effect was reversed by changing the temperature portantly, causes the formation of NO (5, 27). Recent to the opposite direction. This is important, since it has experimental evidence demonstrates that NO is involved been established that temperature directly influences per- in the regulation of permeability (13, 14, 17, 21). Inhibimeability (25). Thus our measurements of P, may be tion of NO synthesis was found to cause an increase in
flow velocity; the flow sensitivity of albumin permeability was 8% increased in P, per millimeter per second increase in velocity. This flow-dependent increase in albumin flux was diminished by the NO synthase inhibitor E-NMMA and restored by L-arginine, the physiological precursor to NO. Moreover, D-NMMA did not influence the permeability. These results suggest that flow modulates coronary venular permeability through the production of NO.
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
FLOW
MODULATES
CORONARY
postcapillary venular permeability in rat and feline intestine in vivo (14, 17). This effect might be attributed to augmented neutrophil adherence when the antiadhesion effect of NO is absent (18), and not a direct effect of NO on the endothelium. In contrast, the increase in albumin permeability induced by substance P was inhibited by the NO synthase inhibitors w-nitro-L-arginine methyl ester and L-NMMA but not by the enantiomer D-NMMA. Moreover, L-arginine reversed the inhibitory effect (13). These latter data corroborate our findings by suggesting that NO may increase permeability of exchange vessels. In our study, L-NMMA caused a decrease in resting permeability and eliminated the flow-dependent permeability. L-Arginine, which is known to reverse the effect of L-NMMA (26), not only restored the flow dependency of permeability but also augmented the resting permeability 2.5fold of control value. This latter finding is puzzling. One possible explanation was that L-arginine influenced transendothelial flux of albumin by altering the charge selectivity of the venular wall. However, the same dose of L-arginine in the absence of L-NMMA did not significantly increase the resting permeability in our control studies. Thus the restoration of resting and flowdependent permeability by L-arginine seems to be attributed to the restoration of NO synthesis following inhibition by L-NMMA. The mechanisms by which NO increases venular permeability are not apparent. Vasodilators that are known to elevate guanosine 3’,5’-cyclic monophosphate (cGMP) in vitro, such as atria1 natriuretic peptide and sodium nitroprusside, and the inhibitor of cGMP breakdown M&B 22948 increased hydraulic conductivity and albumin permeability in frog mesentery capillary (15, 21). It was suggested that these cGMP-dependent agents increase capillary permeability by relaxation of endothelium and/or pericytes, and the subsequent alterations in the dimensions of the endothelial junctions result in higher permeability (21). Thus it is possible that NO, produced by the venular endothelium in response to flow, increases permeability of coronary exchange vessels by a cGMP-dependent mechanism. Such a mechanism is plausible because it is well established that flow (shear stress) initiates the production of NO in arteries and arterioles, which causes vascular relaxation by stimulating the production of cGMP. Alternatively, NO may influence endothelial permeability by production of reactive chemical species. NO reacts with superoxide anion radical to produce peroxynitrite anion, which generates a strong oxidant with reactivity similar to hydroxyl radical (3, 22). Free radicals have been implicated in alteration of endothelial permeability (28). In conclusion, our study demonstrates that flow modulates transvascular exchange of macromolecules in coronary venules by stimulating the production of NO. Further studies need to be carried out to understand the mechanisms of NO-induced permeability changes. We thank Ruth A. Mier for excellent technical assistance. This study was supported by American Heart Association, Texas Affiliate, Grant 91G-024 and by National Heart, Lung, and Blood Institute Grants HL-32788, HL-17669, and HL-21498. Address for renrint reauests: Y. Yuan. Dent. of Medical Phvsiologv.
VENULAR
H645
PERMEABILITY
Texas A & M Univ. Health 1114.
Science Center, College Station,
TX 77843-
Received 24 April 1992; accepted in final form 1 June 1992. REFERENCES 1. Ashikawa, K., H. Kanatsuka, T. Suzuki, and T. Takishima. Phasic blood flow velocity pattern in epimyocardial microvessels in the beating canine left ventricle. Circ. Res. 59: 704-711, 1986. M., and H. Wayland. On-line volume flow rate and 2. Baker, velocity profile measurement for blood in microvessels. lMicrouasc. Res. 7: 131-143, 1974. J. S., T. W. Beckman, J. Chen, P. A. Marshall, 3. Beckman, and B. A. Freeman. Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc. Nutl. Acad. Sci. USA 87: 1620-1624, 1990. 4. Borders, J. L., and H. J. Granger. An optical doppler intravital velocimeter. Microvasc. Res. 27: 117-127, 1984. 5. Buga, G. M., M. E. Gold, J. M. Fukuto, and L. J. Ignarro. Shear stress-induced release of nitric oxide from endothelial cells grown on beads. Hypertension Dallas 17: 187-193, 1991. 6. Caldwell, J. H., G. V. Martin, and J. B. Bassingthwaighte. Intramyocardial permeability-surface area products are proportional to flow: optimality in substrate transport (Abstract). FASEB J. 3: A404, 1989. 7. Crone, C., and D. G. Levitt. Capillary permeability to small System. solutes. In: Handbook of Physiology. The Cardiovascular Microcirculation. Bethesda, MD: Am. Physiol. Sot., 1984, sect. 2, vol. IV, chapt. 10, p. 411-466. F. E., and W. L. Joyner. Modulation of capillary per8. Curry, meability: methods and measurements in individually perfused mammalian and frog microvessels. In: Endothelial Cells. Boca Raton, FL: CRC, 1988, vol. I. p. 3-17. F. E., W. L. Joyner, and J. C. Rutledge. Graded 9. Curry, modulation of frog microvessel permeability to albumin using ionophore A23187. Am. J. Physiol. 258 (Heart Circ. Physiol. 27): H587-H598, 1990. J. M., and T. M. Hollis. Shear stress and aortic 10. DeForrest, histamine synthesis. Am. J. Physiol. 234 (Heart Circ. Physiol. 3): H701-H705, 1978. M. A. Gimbrone, Jr., 11. Dewey, C. F., Jr., S. R. Bussolari, and P. F. Davies. The dynamic response of vascular endothelial cells to fluid shear stress. J. Biomech. Eng. 103: 177-185, 1981. 12. Grega, G. J., S. W. Adamski, and D. E. Dobbins. Physiological and pharmacological evidence for the regulation of permeability. Federation Proc. 45: 96-100, 1986. and S. D. Brain. Evidence that 13. Hughes, S. R., T. J. Williams, endogenous nitric oxide modulates oedema formation induced by substance P. Eur. J. Pharmacol. 191: 481-484, 1990. I. R., B. J. R. Whittle, and N. K. Boughton14. Hutcheson, Smith. Role of nitric oxide in maintaining vascular integrity in endotoxin-induced acute intestinal damage in the rat. Br. J. Pharmacol. 101: 815-820, 1990. V. H., and D. J. Meyer, Jr. Atria1 natriuretic peptide 15. Huxley, (ANP)-induced increase in capillary albumin and water flux. In: Vascular Endothelium in Health and Disease. New York: Plenum, 1988, p. 23-31. and J. M. Tarbell. Endo16. Jo, H., R. 0. Dull, T. M. Hollis, thelial albumin permeability is shear dependent, time dependent, and reversible. Am. J. Physiol. 260 (Heart Circ. Physiol. 29): Hl992-H1996, 1991. P., and D. N. Granger. Nitric oxide modulates mi17. Kubes, crovascular permeability. Am. J. Physiol. 262 (Heart Circ. Physiol. 31): H611-H615, 1992. and D. N. Granger. Nitric oxide: an 18. Kubes, P., M. Suzuki, endogenous modulator of leukocyte adhesion. Proc. Ncztl. Acad. Sci. USA 88: 4651-4655, 1991. Myogenic activity in 19. Kuo, L., M. J. Davis, and W. M. Chilian. isolated subepicardial and subendocardial coronary arterioles. Am. J. Physiol. 255 (Heart Circ. Physiol. 24): Hl558-H1562, 1988. 20. Laine, G. A., and H. J. Granger. Microvascular, interstitial, and lymphatic interactions in normal heart. Am. J. Physiol. 249 (Heart Circ. Physiol. 18): H834-H842, 1985.
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
H646
FLOW
MODULATES
CORONARY
Meyer, D. J,, and V. H. Huxley. Capillary hydraulic conductivity is elevated by cGMP-dependent vasodilators. Circ. Res. 70: 382-391, 1992. M. S., J. M. Hevel, M. A. Marletta, and P. A. 22. Mulligan, Ward. Tissue injury caused by deposition of immune complexes is L-arginine dependent. Proc. Nutl. Acad. Sci. USA 88: 6338-6342, 1991. 23. Overholser, K. A., M. J. Bhatte, and M. H. Laughlin. Modeling the effect of flow heterogeneity on coronary permeabilitysurface area. J. Appl. Physiol. 71: 758-769, 1991. 24. Pilati, C. F. Macromolecular transport in canine coronary microvasculature. Am. J. Physiol. 258 (Heart Circ. Physiol. 27): 21.
H748-H753, 25.
1990.
Rasio, E. A., M. Bendayan, and C. A. Goresky. Effect of temperature change on the permeability of eel rete capillaries. Circ.
Res, 70: 272-284,
1992.
VENULAR 26.
PERMEABILITY
Rees,D. D., R. M. J. Palmer, R. Schulz, H, F. Hodson,and S. Moncada. Characterization of three inhibitors of endothelial nitric oxide synthase in vitro and in vivo. Br. J. Phurmucol. 101: 746-752,
1990.
Rubanyi, G. M., J. C. Romero, and P. M. Vanhoutte. Flow induced release of endothelium-derived relaxing factor. Am. J. Physiol. 250 (Heart Circ. Physiol. 19): H1145-H1149, 1986. 28. Shasby, D. M., S. E. Lind, S. S. Shasby, J. C. Goldsmith, and G. W. Hunninghake. Reversible oxidant-induced increases in albumin transfer across cultured endothelium: alterations in cell shape and calcium homeostasis. Blood 65: 605-614, 1985. 29. Taylor, A. E., and D. N. Granger. Exchange of macromolecules across the microcirculation. In: Handbook of Physiology. The Cardiovascular System. Microcirculation. Bethesda, MD: Am. Physiol. Sot., 1984, sect. 2, vol. IV, chapt. 11, p. 467-520. 27.
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
HARRIS
J. GRANGER,
DAVID
C. ZAWIEJA,
AND WILLIAM
Microcirculation Research Institute and Department of Medical Physiology, Texas A & M University Health Science Center, College Station, Texas 77843-l Yuan, Yuan, Harris J. Granger, David C. Zawieja, and William M. Chilian. Flow modulates coronary venular permeability by a nitric oxide-related mechanism. Am. J. Physiol. 263 (Heart Circ. Physiol. 32): H641-H646, 1992.-This study demonstrates that flow velocity modulates coronary venular permeability to albumin. Apparent permeability coefficients of albumin (P,) were measured in isolated cannulated coronary venules ranging from 30 to 70 pm in diameter. Hydrostatic and oncotic pressures were controlled while the intraluminal flow velocity was varied. P, at an intraluminal hydrostatic pressure of 12 cmH,O and a flow velocity of 7 mm/s was 4.01 t flow velocity to 10 and 13 mm/s 0.53 x 10m6 cm/s. Increasing augmented the permeability by 33 ,t 14 and 48 t 14%, respectively. The nitric oxide synthase inhibitor, NG-monomethyl+ arginine (L-NMMA) ( 10e5 M), decreased baseline P, and abolished the flow-induced permeability changes. Administration of L-arginine (3 x 10m3 M), a physiological precursor of nitric oxide which reverses the effect of L-NMMA, restored the relationship between flow and permeability. From these results we conclude that 1) flow velocity should be considered as a physical force that potentially modulates permeability of venular exchange vessels in the heart and 2) flow modulates coronary venular permeability via the production of nitric oxide. coronary microcirculation; exchange vessels; ability coefficient; NCI-monomethyl-L-arginine;
apparent permeL-arginine
THE HEART is an organ that depends on a continual
supply of oxygen and nutrients for proper function. Adequate delivery of nutrients to the myocytes depends not only on blood flow but also on the microvascular exchange of nutrients. Conventionally, microvascular exchange is thought to be influenced by physical factors, such as hydrostatic and oncotic pressures, and chemical factors, such as inflammatory mediators (12). Yet, such factors cannot fully explain observations showing that increases in flow are often accompanied by increases in the permeability-surface area product (6, 23). Most studies have focused on a change in surface area during alterations in flow but have not eliminated the possibility that flow may have directly influenced permeability. Because the heart is dependent on fatty acids for metabolic fuel and these substrates are normally bound to albumin, it would seem advantageous for an increase in flow, which typically occurs during enhanced myocardial metabolism, to augment microvascular permeability, i.e., facilitate both delivery and exchange. Therefore, we hypothesized that an increase in flow would augment the permeability of coronary exchange vessels. We tested this hypothesis by studying the relationship between venular permeability and flow velocity in isolated
M. CHILIAN 114
coronary venules. Venules were studied because these vessels are important sites for blood-tissue exchange of water and solutes. Accordingly, variations in venular permeability have the potential to alter exchange of water, protein, and nutrients. Because of the well-established link between flow and the production of nitric oxide (NO) (5, 27), and our finding that permeability and flow are also associated, we further speculated that NO is involved in flow-induced permeability changes. MATERIALS General
AND METHODS
Preparation
Pigs weighing 9-13 kg were sedated with ketamine (2.5 mg/kg im) and Rompun (2.25 mg/kg im), anesthetized with Nembutal (25 mg/kg iv), and heparinized (250 units/kg iv). Following a tracheotomy and intubation, the animals were ventilated. A left thoractomy was performed, and the heart was electrically fibrillated, excised, and placed in 4°C physiological saline. The coronary sinus was cannulated, and 3 ml india ink-gelatin-physiological salt solution were infused to clearly define venular microvessels. This solution was prepared by adding 0.2 ml of nondialyzed india ink (Koh-I-Noor, Bloomsbury, NJ) and 0.35 g of porcine skin gelatin to IO ml of warm physiological salt solution and filtering through P8 filter paper (Fisher Scientific, Pittsburgh, PA). This perfusate does not affect microvascular endothelial reactivity (19). Solutions
and Perfusates
The albumin-physiological salt solution (APSS) used for venule dissection had the following composition (in mM): 145.0 NaCl, 4.7 KCl, 2.0 CaCl,, 1.17 MgSO,, 1.2 NaH,PO,, 5.0 glucose, 2.0 pyruvate, 0.02 EDTA, and 3.0 3-(N-morpholino)propanesulfonic acid (MOPS) buffer. After adding 1% bovine serum albumin, the solution was buffered to a pH of 7.40 at 4°C and then filtered through a Millex-PF O.&pm filter unit (Millipore, Bedford, MA). The APSS used to perfuse the vessels had the same composition as mentioned above but was buffered to a pH of 7.40 at 37°C and filtered. The test solute for permeability measurements was fluorescein isothiocyanate-bovine albumin (FITC-albumin) (Sigma, St. Louis, MO). The test solutions were prepared by adding 5 mg FITC-albumin to 100 ml APSS. Isolated
Venule
Preparation
The first step of the dissection procedure was visualization of a suitable venule (length, 0.8-1.2 mm; diameter, 30-70 pm) with a Nikon dissecting microscope. The venule and a small amount of myocardium were excised from the surrounding tissue and then moved to a chamber containing APSS (4°C) for further dissection. Dumont forceps and iridectomy scissors were used to remove the tissue surrounding the venule until the desired
0363-6135/92 $2.00 Copyright 0 1992 the American Physiological Society Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
H641
H642
FLOW
MODI
‘LATES
CORONARY
length of venule was cleared of cardiac tissue. The exposed portion of the vessel was then cut and transferred to the cannulating chamber, which was mounted on a Zeiss Axiovert microscope and gradually warmed to 37°C. The isolated venule was cannulated on both the inflow and outflow ends with two micropipettes (inflow pipette tip: OD 40 pm, ID 30 pm; outflow pipette tip: OD 40 pm, ID 20 pm) and secured with 11-O suture (Alcon, Fort Worth, TX). A third pipette (tip: OD 20pm, ID 16 pm) was inserted into the inflow pipette. The vessels were perfused with either APSS through the outer inflow pipette or APSS containing fluorescently labeled test solutes through the inner inflow pipette (Fig. 1). The cannulating micropipettes were connected to three reservoirs so that intraluminal pressure and flow could be adjusted independently by simultaneously changing the heights of inflow and outflow reservoirs in equal increments. For instance, setting the inflow reservoir at 20 cmH,O and the outflow reservoir at 0 cmH,O produced an intraluminal pressure of 12 cmH,O and an intraluminal flow velocity of 7 mm/s. (Intraluminal pressure was slightly higher than the average of inflow and outflow pressures because of the difference of resistance between the tips of the inflow and outflow pipettes.) Simultaneously raising the inflow reservoirs to 30 cmH,O and decreasing outflow reservoir to -10 cmH,O increased intraluminal flow velocity to 13 mm/s while maintaining a relatively constant intraluminal pressure (intraluminal pressures increased slightly to 15 cmH,O). Because the apparent permeability is influenced by pressure, possible effects of these small changes in pressure during increases in flow velocity were corrected by factoring out the pressuredependent flux of protein. To accomplish this, the permeability to albumin was measured at constant flow during elevation of inflow reservoirs 20 cmH20 APSS FlTC tflow reservoir 0 cmH20
VENULAR
pressure to 12 and 15 cmH,O, which was produced by elevating both the inflow and outflow reservoirs. These values were then subtracted from the apparent permeability observed at the two high flow rates. The image of the vessel (x20) was projected onto a siliconintensified target camera and displayed on a high-resolution monochromatic video monitor. The vessel diameter was measured on the monitor with a calibrated video caliper. The intraluminal pressure was measured through a pressure transducer connected to the inner inflow pipette when the pipette tip was advanced into the venular lumen. The intraluminal flow velocity was measured with an optical Doppler velocimeter (Microcirculation Research Institute, Texas A & M University) (4) by perfusing the vessel with a 1% red blood cell suspension. The preparation was discarded if leakage of the fluorochromes occurred. Measurement
II
l-----JL,
1
‘i-r’~e
7
inner pipette
outer pipette
outflow pipette
Fig. 1. Schematic diagram showing pipettes and reservoirs that control intraluminal flow and pressure of cannulated venule. Venule was cannulated with 2 inflow pipettes (outer and inner) and an outflow pipette. The 2 inflow reservoirs connected to inner and outer pipettes were mounted toge ther so that they could be raised or lowered simultaneously. Increasing inflow reservoirs and decreasing outflow reservoir in equal increments increased intraluminal flow velocity while intraluminal pressure was maintained at a relative constant level . Simultaneously increasing inflow and outflow reservoirs in equal increments increased intraluminal pressure wi thout significant changes in flow velocity. This system enables precise measurements of venular permeability under conditions in which forces that govern permeability and factors that influence permeability are controlled. APSS, albumin-physiological salt solution; FITC, fluorescein isothiocyanate.
of Venular
Permeability
The permeability of the vessel was measured by the ratio of the transmural flux of fluorescently labeled molecules per unit surface area per unit time to the transmural difference of solute concentration, as developed by Huxley and Curry (8). With the use of an optical window of a video photometer (Microcirculation Research Institute, Texas A & M University) positioned over the venules and adjacent space on the monitor, the fluorescent intensity from the window Twas measured. In each measurement, the isolated venule was first perfused with APSS through the outer inflow pipette to establish a baseline intensity. The venular lumen was then rapidly filled with fluorochromes by switching the perfusion to the inner inflow pipette, which produced a step increase and subsequently a gradual increase in the fluorescent intensity. There was a step decrease of intensity when the fluorochromes were washed out by switching the perfusion to the outer inflow pipette. The permeability coefficient to albumin (P,) was calculated using the equation P = (l/AI,)(dI,ldt),(r/2), where AI, is the step increase in fluorescent intensity, (dI,ldt), is the initial rate of increase in intensity as solutes diffuse out of the vessel and into the extravascular space, and r is the venular radius. Measurements were repeated at l- to 2-min intervals. Experimental
APSS
PERMEABILITY
Protocols
Effect of flow on venular permeability. The baseline P, was measured in 12 isolated coronary venules (12 vessels from 12 pigs) under an intraluminal pressure of 12 cmH,O and flow velocity of 7 mm/s. The relationship between P, and flow velocity was tested by measuring P, at a series of intraluminal velocities. To produce a series of flow velocities while maintaining the intraluminal pressure relatively constant, the heights of inflow and outflow reservoirs were adjusted as follows: 1) inflow 20 cmH,O, outflow 0 cmH20; 2) inflow 25 cmH,O, outflow -5 cmH,O; and 3) inflow 30 cmH,O, outflow -10 cmH,O, producing flow velocities of 7, IO, and 13 mm/s, respectively. These velocities are in the range of in vivo measurements of flow velocities in small coronary venules (I). In four preparations, a 1% red blood cell suspension was introduced into the lumen, and flow velocity at various pressures was measured with an optical Doppler velocimeter. Effect of NO on flow-related venularpermeability. The role of NO in flow-dependent permeability changes was studied by comparing the effect of flow on venular permeability during control conditions (n = 12) and after administration of NO synthase inhibitor NG-monomethyl+arginine (L-NMMA; low5 M) in the suffusion bath for 20-30 min in 8 of the above 12 vessels. To establish that the effects of L-NMMA were specific, P, was then measured in the presence of L-arginine (3 X 10e3 M; n = 3), which is reported to reverse the effects of the antagonist
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
FLOW
MODULATES
CORONARY
VENULAR
(26). To eliminate the possibility that the effects of L-NMMA on permeability were mediated by its charge, the same concentration of NG-monomethyl-D-arginine (D-NMMA), which is a biologically inactive isomer of L-NMMA, was applied for comparison. Data
H643
PERMEABILITY Y
Control
v
L-NMMA
Y
L-arginine
Analysis
P, was measured two or three times at each flow velocity and averaged. The mean P, at each flow level was normalized to the control P, at 7 mm/s, i.e., (P, at certain flow/P, at 7 mm/s) X 100, to obtain the relationship between intraluminal flow velocity and percentage changes of P,. Regression analysis was applied to examine the statistical significance. The differences were considered significant when P < 0.05. All the data were presented as means t SE. RESULTS Effect of Flow on Venular
Permeability
Under control conditions, the P, obtained from 12 vessels was 4.01 2 0.53 X lOA6 cm/s at an intraluminal pressure of 12 cmH20 and a flow velocity of 7 mm/s. Increases in flow velocity to 10 and 13 mm/s augmented the uncorrected permeability by 48 t 14 and 71 t 14%, respectively (Fig. 2A, dashed line). Intraluminal pressure also increased modestly during the increase in flow (2 and 3 cmHzO, respectively, from the control pressure of 200
50 I
0 5
, 7
Flow
I
1 9
Velocity
I
1
I
11
I
’
13
1
15
(mmkec)
Fig. 3. Albumin permeability in isolated coronary venules increased 48% as intraluminal flow increased 6 mm/s. Effect of flow on permeability was abolished by treating same venules with NG-monomethylL-arginine (L-NMMA; lo+ M). L-Arginine (10v3 M) in the presence of L-NMMA restored the flow-induced permeability change. All data have been corrected for effects of pressure (refer to Fig. 2 for details).
A cu n
180
E 6
140
z b0 F
120 160
2
80
! 6
I 8
Flow
I 10
Velocity
I 12
1 14
(mmkec)
12 cmH20). To eliminate the potential influence of these small intraluminal pressure changes on P,, we measured P, at the three pressures during constant flow to formulate a relationship between P, and pressure (Fig. 2B). The effects of pressure were then subtracted to provide the pressure-independent relationships between flow velocity and P, (Fig. 2A, solid line). Importantly, the corrected P, still increased significantly as a function of flow velocity: P, increased 33 and 48% as intraluminal velocity increased 3 and 6 mm/s, respectively, indicating that venular permeability is flow dependent. Effect of NO on Flow-Related
Q e
180
0 L
160
I
80
i 11
I 12
Intraluminal
I 13
I 14
Pressure
'
I
15
(cmH
m
I
16
*O)
Fig. 2. A: apparent permeability coefficient of albumin (Pa) increased as a function of flow velocity in isolated coronary venules. Relationships between P, and flow velocity before (dashed line) and after correction (solid line) are shown. Correction is to factor out effects of slight increases in intraluminal pressure that occurred during increases in flow velocity (intraluminal pressures measured at flow velocities of 7, 10, and 13 mm/s were 12, 14, and 15 cmH,O, respectively). B: relationship between P, and pressure ranging from 12 to 15 cmH20, which was obtained at constant flow by elevating both inflow and outflow pressures by identical amounts. This slope serves as correction factor to subtract effect of small pressure changes in curve of P, vs. flow shown in A.
Venular
Permeability
After L-NMMA treatment (10u5 M; n = 8), the baseline P, decreased from the control value of 4.01 t 0.53 to 3.15 t 0.58 x lo+ cm/s, and the effect of flow on permeability was abolished (Fig. 3). Two-way analysis of variance showed a significant difference between the two curves (P < 0.05). L-Arginine (3 x 10e3 M; n = 3) not only restored the relation between flow and permeability but increased P, 2-3 times over control. D-NMMA (low5 M) did not alter the resting permeability or the flow-induced increase in permeability. Under control conditions, L-arginine alone did not influence permeability. DISCUSSION
This study demonstrates that flow, via the production of NO, is capable of modulating permeability of coronary venules. The apparent permeability coefficient of albumin was measured in isolated coronary venules ranging from 30 to 70 ,um in diameter, in which hydrostatic and oncotic pressure were controlled, but intraluminal flow velocity was varied. P, increased as a function of
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
H644
FLOW
MODULATES
CORONARY
VENULAR
PERMEABILITY
greater than those obtained in isolated frog vessels (9) partially because of the differences between temperatures in the two studies (37 vs. 15°C). On the basis of these arguments, it seems unlikely that the vessels were functionally damaged during isolation and perfusion procedures. It is our belief that the relatively high P, was not an artifact related to damage, but rather reflects the highly permeable nature of the coronary exchange vessels. As stated above, the relatively high permeability of Methodological Considerations coronary venules most likely reflects a physiological charThe apparent permeability coefficient of albumin was acteristic of coronary exchange vessels. Laine and Granger (20) showed that the transvascular flux of fluid measured from the ratio of transvenular flux of fluorescent-labeled solutes per unit surface area per unit time to and protein in the myocardium was at least one order of the transmural difference of solute concentration. This magnitude higher than in skin or skeletal muscles. This analysis assumes that albumin crosses the venular wall high efflux was not explained fully by the larger exchange exclusively by diffusion. Because macromolecular ex- surface area in myocardium because the lymph-to-plasma travasation is also coupled to water filtration, the appar- ratios for total plasma protein (0.8) and for albumin (0.9) ent permeability coefficient of albumin obtained in this were also higher than those in the other tissues such as study is an overestimation of the true permeability of skeletal muscle (29). Pilati (24) found that coronary vascoronary venules. Because we took precautions to stabi- cular reflection coefficient for albumin was relatively low lize the transmural pressure and account for pressure- (0.59 t 0.05) when compared with that of skeletal muscle dependent contributions, the changes in apparent P, re- and mesentery (7). An explanation for the comparatively high coronary permeability is that the heart has higher flect alterations in albumin permeability. We would like to point out that our measurements of metabolic demands than other tissues, and the relatively permeability reflected greater transmural flux and not high metabolism necessitates a rather rapid exchange of nutrients and substrates in the coronary exchange altered vascular filling. This is important to highlight because in the formula to calculate permeability, a change vessels. The flow velocities employed in this study were 7-13 in AI, (step change in fluorescence associated with filling) and/or change in (dIJdt). (albumin flux) alters the cal- mm/s, which are within the normal physiological range of culated permeability. Filling of the vessels (AI,) was not coronary venular flow velocity (1). This range of flow changed; rather, only the flux values (dIJdt), were velocity applied in the vessels ranging from 30 to 70 pm in changed by the interventions. Moreover, venular diame- diameter produced a shear stress in a range of 3.85-16.68 ter did not change during the experimental maneuvers. dyn/cm2. Shear stress (y) was calculated by y = ps, where Thus we can state with conviction that we have studied p is viscosity of the testing solution, which was measured as 0.77 CP in our study, and s is wall shear rate calculated changes in venular permeability as opposed to artifacts associated with altered filling or diameter changes. bY s = 5 x centerline velocity/vessel diameter (2). One P, measured at intraluminal pressure of 12 cmHZO and concern is that alteration of intraluminal velocity, and flow velocity of 7 mm/s was 4.01 t 0.53 X lo+ cm/s, thus shear stress, may have induced changes in vessel which is one order of magnitude larger than that in frog diameter, which would compromise permeability meamesenteric capillaries (4.7 t 1.1 X 10e7 cm/s at 8.1 t 3.5 surements. We did not find any alterations in vascular cmH20) (9). We would argue that this high baseline per- caliber during variations in intraluminal flow velocity. meability was not caused by vessel damage during isola- Importantly, the coronary venules we studied consisted tion because of several observations and facts. 1) Mea- primarily of a single layer of endothelial cells and several surement of P, in isolated hamster cheek pouch venules pericytes; thus they most likely had little active tone. using similar techniques was 2.1 x 10e6 cm/s, which was Therefore, changes in permeability were not due to comparable to that obtained from in situ preparations changes in venular diameter, but were due to the effects of (1.3 x 10m6 cmHZO) (8). 2) P, was stable throughout the flow velocity. experiments and was reversibly affected by the interventions, suggesting that the barrier function of the vessel Flow Modulates Coronary Venular Permeability wall was not severely damaged, and was not progressively A recent study indicated that albumin permeability of deteriorating. 3) Administration of indomethacin, an bovine aortic endothelial monolayer is shear dependent anti-inflammatory agent, did not alter the P, of the ves- and reversible (16). It is difficult to extrapolate results sels, which implies that the experimental procedures did from cultured aortic endothelium to factors that modunot cause prostanoid production associated with damage. late exchange processes in the microvasculature. Our 4) P, was measured at a normal body temperature (36- study suggests a role of flow velocity in regulation of 37°C). Our preliminary experiments showed that in- venular permeability. Shear stress influences physical and biochemical charcreases in temperature from 15 to 36-37°C increased P, two- to threefold in isolated porcine coronary venules, acteristics of endothelial cells (10, 11, 16) and, more imand this effect was reversed by changing the temperature portantly, causes the formation of NO (5, 27). Recent to the opposite direction. This is important, since it has experimental evidence demonstrates that NO is involved been established that temperature directly influences per- in the regulation of permeability (13, 14, 17, 21). Inhibimeability (25). Thus our measurements of P, may be tion of NO synthesis was found to cause an increase in
flow velocity; the flow sensitivity of albumin permeability was 8% increased in P, per millimeter per second increase in velocity. This flow-dependent increase in albumin flux was diminished by the NO synthase inhibitor E-NMMA and restored by L-arginine, the physiological precursor to NO. Moreover, D-NMMA did not influence the permeability. These results suggest that flow modulates coronary venular permeability through the production of NO.
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
FLOW
MODULATES
CORONARY
postcapillary venular permeability in rat and feline intestine in vivo (14, 17). This effect might be attributed to augmented neutrophil adherence when the antiadhesion effect of NO is absent (18), and not a direct effect of NO on the endothelium. In contrast, the increase in albumin permeability induced by substance P was inhibited by the NO synthase inhibitors w-nitro-L-arginine methyl ester and L-NMMA but not by the enantiomer D-NMMA. Moreover, L-arginine reversed the inhibitory effect (13). These latter data corroborate our findings by suggesting that NO may increase permeability of exchange vessels. In our study, L-NMMA caused a decrease in resting permeability and eliminated the flow-dependent permeability. L-Arginine, which is known to reverse the effect of L-NMMA (26), not only restored the flow dependency of permeability but also augmented the resting permeability 2.5fold of control value. This latter finding is puzzling. One possible explanation was that L-arginine influenced transendothelial flux of albumin by altering the charge selectivity of the venular wall. However, the same dose of L-arginine in the absence of L-NMMA did not significantly increase the resting permeability in our control studies. Thus the restoration of resting and flowdependent permeability by L-arginine seems to be attributed to the restoration of NO synthesis following inhibition by L-NMMA. The mechanisms by which NO increases venular permeability are not apparent. Vasodilators that are known to elevate guanosine 3’,5’-cyclic monophosphate (cGMP) in vitro, such as atria1 natriuretic peptide and sodium nitroprusside, and the inhibitor of cGMP breakdown M&B 22948 increased hydraulic conductivity and albumin permeability in frog mesentery capillary (15, 21). It was suggested that these cGMP-dependent agents increase capillary permeability by relaxation of endothelium and/or pericytes, and the subsequent alterations in the dimensions of the endothelial junctions result in higher permeability (21). Thus it is possible that NO, produced by the venular endothelium in response to flow, increases permeability of coronary exchange vessels by a cGMP-dependent mechanism. Such a mechanism is plausible because it is well established that flow (shear stress) initiates the production of NO in arteries and arterioles, which causes vascular relaxation by stimulating the production of cGMP. Alternatively, NO may influence endothelial permeability by production of reactive chemical species. NO reacts with superoxide anion radical to produce peroxynitrite anion, which generates a strong oxidant with reactivity similar to hydroxyl radical (3, 22). Free radicals have been implicated in alteration of endothelial permeability (28). In conclusion, our study demonstrates that flow modulates transvascular exchange of macromolecules in coronary venules by stimulating the production of NO. Further studies need to be carried out to understand the mechanisms of NO-induced permeability changes. We thank Ruth A. Mier for excellent technical assistance. This study was supported by American Heart Association, Texas Affiliate, Grant 91G-024 and by National Heart, Lung, and Blood Institute Grants HL-32788, HL-17669, and HL-21498. Address for renrint reauests: Y. Yuan. Dent. of Medical Phvsiologv.
VENULAR
H645
PERMEABILITY
Texas A & M Univ. Health 1114.
Science Center, College Station,
TX 77843-
Received 24 April 1992; accepted in final form 1 June 1992. REFERENCES 1. Ashikawa, K., H. Kanatsuka, T. Suzuki, and T. Takishima. Phasic blood flow velocity pattern in epimyocardial microvessels in the beating canine left ventricle. Circ. Res. 59: 704-711, 1986. M., and H. Wayland. On-line volume flow rate and 2. Baker, velocity profile measurement for blood in microvessels. lMicrouasc. Res. 7: 131-143, 1974. J. S., T. W. Beckman, J. Chen, P. A. Marshall, 3. Beckman, and B. A. Freeman. Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc. Nutl. Acad. Sci. USA 87: 1620-1624, 1990. 4. Borders, J. L., and H. J. Granger. An optical doppler intravital velocimeter. Microvasc. Res. 27: 117-127, 1984. 5. Buga, G. M., M. E. Gold, J. M. Fukuto, and L. J. Ignarro. Shear stress-induced release of nitric oxide from endothelial cells grown on beads. Hypertension Dallas 17: 187-193, 1991. 6. Caldwell, J. H., G. V. Martin, and J. B. Bassingthwaighte. Intramyocardial permeability-surface area products are proportional to flow: optimality in substrate transport (Abstract). FASEB J. 3: A404, 1989. 7. Crone, C., and D. G. Levitt. Capillary permeability to small System. solutes. In: Handbook of Physiology. The Cardiovascular Microcirculation. Bethesda, MD: Am. Physiol. Sot., 1984, sect. 2, vol. IV, chapt. 10, p. 411-466. F. E., and W. L. Joyner. Modulation of capillary per8. Curry, meability: methods and measurements in individually perfused mammalian and frog microvessels. In: Endothelial Cells. Boca Raton, FL: CRC, 1988, vol. I. p. 3-17. F. E., W. L. Joyner, and J. C. Rutledge. Graded 9. Curry, modulation of frog microvessel permeability to albumin using ionophore A23187. Am. J. Physiol. 258 (Heart Circ. Physiol. 27): H587-H598, 1990. J. M., and T. M. Hollis. Shear stress and aortic 10. DeForrest, histamine synthesis. Am. J. Physiol. 234 (Heart Circ. Physiol. 3): H701-H705, 1978. M. A. Gimbrone, Jr., 11. Dewey, C. F., Jr., S. R. Bussolari, and P. F. Davies. The dynamic response of vascular endothelial cells to fluid shear stress. J. Biomech. Eng. 103: 177-185, 1981. 12. Grega, G. J., S. W. Adamski, and D. E. Dobbins. Physiological and pharmacological evidence for the regulation of permeability. Federation Proc. 45: 96-100, 1986. and S. D. Brain. Evidence that 13. Hughes, S. R., T. J. Williams, endogenous nitric oxide modulates oedema formation induced by substance P. Eur. J. Pharmacol. 191: 481-484, 1990. I. R., B. J. R. Whittle, and N. K. Boughton14. Hutcheson, Smith. Role of nitric oxide in maintaining vascular integrity in endotoxin-induced acute intestinal damage in the rat. Br. J. Pharmacol. 101: 815-820, 1990. V. H., and D. J. Meyer, Jr. Atria1 natriuretic peptide 15. Huxley, (ANP)-induced increase in capillary albumin and water flux. In: Vascular Endothelium in Health and Disease. New York: Plenum, 1988, p. 23-31. and J. M. Tarbell. Endo16. Jo, H., R. 0. Dull, T. M. Hollis, thelial albumin permeability is shear dependent, time dependent, and reversible. Am. J. Physiol. 260 (Heart Circ. Physiol. 29): Hl992-H1996, 1991. P., and D. N. Granger. Nitric oxide modulates mi17. Kubes, crovascular permeability. Am. J. Physiol. 262 (Heart Circ. Physiol. 31): H611-H615, 1992. and D. N. Granger. Nitric oxide: an 18. Kubes, P., M. Suzuki, endogenous modulator of leukocyte adhesion. Proc. Ncztl. Acad. Sci. USA 88: 4651-4655, 1991. Myogenic activity in 19. Kuo, L., M. J. Davis, and W. M. Chilian. isolated subepicardial and subendocardial coronary arterioles. Am. J. Physiol. 255 (Heart Circ. Physiol. 24): Hl558-H1562, 1988. 20. Laine, G. A., and H. J. Granger. Microvascular, interstitial, and lymphatic interactions in normal heart. Am. J. Physiol. 249 (Heart Circ. Physiol. 18): H834-H842, 1985.
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
H646
FLOW
MODULATES
CORONARY
Meyer, D. J,, and V. H. Huxley. Capillary hydraulic conductivity is elevated by cGMP-dependent vasodilators. Circ. Res. 70: 382-391, 1992. M. S., J. M. Hevel, M. A. Marletta, and P. A. 22. Mulligan, Ward. Tissue injury caused by deposition of immune complexes is L-arginine dependent. Proc. Nutl. Acad. Sci. USA 88: 6338-6342, 1991. 23. Overholser, K. A., M. J. Bhatte, and M. H. Laughlin. Modeling the effect of flow heterogeneity on coronary permeabilitysurface area. J. Appl. Physiol. 71: 758-769, 1991. 24. Pilati, C. F. Macromolecular transport in canine coronary microvasculature. Am. J. Physiol. 258 (Heart Circ. Physiol. 27): 21.
H748-H753, 25.
1990.
Rasio, E. A., M. Bendayan, and C. A. Goresky. Effect of temperature change on the permeability of eel rete capillaries. Circ.
Res, 70: 272-284,
1992.
VENULAR 26.
PERMEABILITY
Rees,D. D., R. M. J. Palmer, R. Schulz, H, F. Hodson,and S. Moncada. Characterization of three inhibitors of endothelial nitric oxide synthase in vitro and in vivo. Br. J. Phurmucol. 101: 746-752,
1990.
Rubanyi, G. M., J. C. Romero, and P. M. Vanhoutte. Flow induced release of endothelium-derived relaxing factor. Am. J. Physiol. 250 (Heart Circ. Physiol. 19): H1145-H1149, 1986. 28. Shasby, D. M., S. E. Lind, S. S. Shasby, J. C. Goldsmith, and G. W. Hunninghake. Reversible oxidant-induced increases in albumin transfer across cultured endothelium: alterations in cell shape and calcium homeostasis. Blood 65: 605-614, 1985. 29. Taylor, A. E., and D. N. Granger. Exchange of macromolecules across the microcirculation. In: Handbook of Physiology. The Cardiovascular System. Microcirculation. Bethesda, MD: Am. Physiol. Sot., 1984, sect. 2, vol. IV, chapt. 11, p. 467-520. 27.
Downloaded from www.physiology.org/journal/ajpheart at Univ of Cambridge (131.111.164.128) on March 18, 2019.
