Despite the small number of leukocytes relative to erythrocytes in the circulation, leukocytes contribute significantly to organ blood flow resistance. The present study was designed to investigate whether interactions between leukocytes and erythrocytes affect the pressure-flow relationship in a hemodynamically isolated rat gracilis muscle. At constant arterial flow rate, arterial pressure was increased significantly when relatively few physiological counts of leukocytes were added to a suspension containing erythrocytes at physiological hematocrits. However, the arterial pressure after perfusion of similar numbers of isolated leukocytes without erythrocytes was only slightly increased. An increase in resistance was also observed when leukocytes were replaced with 6-μm microspheres. We propose a new mechanism for increasing the hemodynamic resistance that involves hydrodynamic interactions between leukocytes and erythrocytes. In the presence of larger and less deformable leukocytes, erythrocytes move through capillaries more slowly than without leukocytes. Therefore erythrocytes are displaced from their axial positions. Slowing and radial displacement of erythrocytes serve to increase the relative apparent viscosity attributable to erythrocytes, thereby causing a significant elevation of organ blood flow resistance.
- erythrocyte velocity
- skeletal muscle blood flow
- capillary network resistance
a more detailed understanding of whole organ perfusion is fundamental for the analysis of many physiological problems as well as several cardiovascular conditions such as fever (35), stroke, myocardial infarction, and hypertension. The rheological properties of blood (8, 9) have been investigated as a key determinant of the whole organ pressure-flow relationship, and emphasis has been placed on plasma viscosity (31) and erythrocyte properties (32). Because leukocytes constitute <1% of the total number of blood cells in the circulation, the specific contribution of this small fraction to whole organ hemodynamics has traditionally been ignored. However, recent in vivo studies have suggested that leukocytes may have a disproportionate effect on the local blood flow (29). Thus, in addition to their basic role in responding to immunological and inflammatory stimuli, the relatively few circulating leukocytes may play an important rheological role in determining the level of organ blood flow.
Several mechanisms have been proposed to describe the effects of leukocytes on local hemodynamics in skeletal muscle. Circulating leukocytes may become activated by a variety of mechanisms and may release substances that promote vasoconstriction, especially in the presence of atherosclerotic lesions (18, 23). The relatively high cytoplasmic stiffness of the leukocyte may lead to plugging either at the entrance to (4, 10, 34) or within (13-15) the capillaries. Capillary plugging per se may have only a small effect on the whole organ blood flow resistance, however, unless large numbers of leukocytes become entrapped within the microcirculation (10, 15, 34). Leukocyte entrapment within the capillaries may become more important if the capillary endothelium shows the projection of pseudopod structures into the vessel lumen (26), since the relatively rigid mechanical properties of pseudopods may further restrict leukocyte passage through a capillary (19). Leukocytes adhering in postcapillary venules may also restrict organ blood flow by inducing vasoconstriction in neighboring arterioles (22, 25, 37). Organ blood flow resistance may also be increased by hydrodynamic interactions between leukocytes and erythrocytes in the microcirculation (36). Local hemodynamics may be affected by cell-cell collisions during flow through small vessels (33). Furthermore, leukocyte margination in postcapillary venules may be enhanced in the presence of erythrocytes (27), leading to leukocyte-endothelial cell adhesive interactions that may partially obstruct local blood flow (17).
The present study was designed to investigate whether the interaction between leukocytes and erythrocytes leads to a disturbance of the flow characteristics of erythrocytes, thereby causing an elevation of hemodynamic resistance. The pressure-flow relationship in the hemodynamically isolated rat gracilis muscle was examined during perfusion with isolated leukocytes, pure erythrocytes, or a mixture of leukocytes and erythrocytes. Two principal questions were addressed. At a given flow rate, does the sum of individual additional pressure drops induced by isolated leukocytes and pure erythrocytes equal the additional pressure drop measured when both leukocytes and erythrocytes are present? Can similar shifts in the pressure-flow relationships be induced by biologically inert microspheres of comparable dimensions, indicating that an increase in hemodynamic resistance at each flow rate depends primarily on hydrodynamic interactions between leukocytes and erythrocytes?
Animal care during all experiments was approved by the University of California, San Diego, Animal Subjects Committee in accordance with National Institutes of Health guidelines. Male Wistar rats, 280–350 g, were anesthetized by intraperitoneal injection of pentobarbital sodium (40–50 mg/kg body wt, Nembutal, Abbott Laboratories, N. Chicago, IL). Polyethylene catheters (PE-50, 0.58 mm ID, Clay Adams, Parsippany, NJ) were inserted into the left femoral artery and vein to monitor central blood pressure and to deliver additional anesthesia and fluids, respectively. The experimental animal rested on a custom-built Lucite stage that was warmed to maintain body temperature.
Cell-free plasma substitute.
A cell-free suspension (CFS) served as the basic medium for perfusion and as the suspension buffer for blood cells and microspheres. Bovine serum albumin (5%, fraction V powder, Sigma Chemical, St. Louis, MO) was dissolved in a physiological electrolyte solution, Plasma-Lyte A (Baxter Healthcare, Deerfield, IL). Heparin sodium (10 IU/ml, Elkins-Sinn, Cherry Hill, NJ) was added to prevent coagulation.
Blood cell suspensions.
In four experiments, blood from male Wistar donor rats was used to prepare three blood cell suspensions: an isolated leukocyte suspension, a pure erythrocyte suspension, and a suspension containing both leukocytes and erythrocytes.
To prepare the mixed leukocyte-erythrocyte suspension (LES), heparinized donor rat blood was centrifuged for 10 min at 800g. The plasma supernatant was discarded. The remaining buffy coat and erythrocyte layers were resuspended in the CFS to obtain a hematocrit of 40 ± 3% (mean ± SD) and a leukocyte concentration of 5.1 ± 1.3 × 106 leukocytes/ml. The LES contained 84.4 ± 6.1% lymphocytes, 13.5 ± 4.5% polymorphonuclear leukocytes, 2.1 ± 1.0% monocytes, and <1% eosinophils and basophils. These concentrations were not significantly different from those in donor blood (P≥ 0.05, paired t-test).
The isolated leukocyte syspension (LS) was prepared by separating leukocytes from donor rat blood according to a method modified from that of Braide and Bjürsten (5). Briefly, Percoll (Sigma) and hyperosmotic phosphate-buffered saline (PBS, 400 mosM) were used to prepare two suspensions with osmolarity of 400 mosM and densities of 1.0875 mg/ml (solution A) and 1.1025 mg/ml (solution B).Solution A was layered on top of an equal volume of solution B in a centrifuge tube. Heparinized (10 IU/ml) donor rat blood was diluted with an equal volume of PBS (400 mosM). The blood-PBS mixture was layered on top of solution A, and the tube was centrifuged for 30 min at 800g. The plasma supernatant was discarded. The leukocyte bands were collected and diluted with an equal volume of PBS (400 mosM). The leukocyte-PBS mixture was centrifuged for 5 min at 1,000 g. The supernatant was discarded, and the leukocyte pellet was resuspended in the CFS to obtain a total leukocyte concentration the same as that in the LES. The LS contained 86.4 ± 5.4% lymphocytes, 10.9 ± 4.9% polymorphonuclear leukocytes, 2.2 ± 0.9% monocytes, and <1% eosinophils and basophils. These concentrations were not significantly different from either LES or donor blood (P ≥ 0.05, pairedt-test).
After removal of the leukocyte bands, the pure erythrocyte suspension (ES) was prepared by removing the erythrocyte fraction from the bottom of the centrifuge tube. The erythrocytes were diluted in the CFS to obtain a hematocrit which was the same as that in the LES. The total leukocyte concentration in the ES was 131 ± 90 × 103 leukocytes/ml (n = 4).
In six experiments, polystyrene microspheres with a nominal diameter of 6.0 μm (Polysciences, Warrington, PA) were used to simulate the particle flow dynamics of leukocytes. Microspheres were suspended in the CFS at concentrations of ∼7.5 × 106 particles/ml to create an isolated microsphere suspension (MS).
Heparinized donor blood from male Wistar rats was centrifuged for 10 min at 800 g, and the packed erythrocyte layer was drawn from the bottom of the tube. ES was prepared by resuspending the packed erythrocytes in the CFS to obtain hematocrits of 11, 26, or 35% and total leukocyte concentrations of <0.1 × 106 leukocytes/ml. Two experiments were performed at each hematocrit.
Mixed microsphere-erythrocyte suspensions (MES) were prepared by suspending microspheres and donor erythrocytes in the CFS. The particle concentrations were adjusted so that the hematocrit and the microsphere concentration were the same as those in ES and a suspension containing microspheres but no erythrocytes (MS), respectively.
Gracilis muscle preparation.
The objective of the surgical procedure was to hemodynamically isolate the right gracilis muscle from the animal’s central circulation. The exposed muscle was superfused continuously with warmed Krebs-Henseleit buffer bubbled with 95% N2-5% CO2 to maintain tissue viability. After all visible vessels communicating between the gracilis muscle and the surrounding tissue were ligated, the femoral artery and vein were cannulated near the abdominal wall. The arterial catheter was connected to a precise piston-type perfusion pump (28), and the venous catheter was allowed to drain at atmospheric pressure. CFS was infused at a steady pressure of 50 mmHg for a 10-min equilibration period, and observation of the muscle revealed no significant leakage of perfusate into the surrounding tissue. At the end of each experiment, a carbon ink suspension (Pelikan, Germany) was infused to mark perfused areas of the gracilis muscle. The ink-perfused area of the muscle was excised and weighed.
The steady-state pressure-flow relationship was measured by stepping the input arterial flow rate down in 6–10 steps from ∼1.0 to 0.0 ml/min in time intervals of 15–20 s. At each flow step, the corresponding input arterial pressure was measured.
In four experiments, suspensions were perfused through the isolated gracilis muscle in the following order: CFS, LS, ES, and LES. In six separate experiments, the leukocytes were replaced with microspheres, and the following suspensions were perfused: CFS, MS, ES, and MES. Preliminary studies showed that the order of perfusion did not affect the existence of a leukocyte-mediated increase in perfusion pressure at each flow rate.
Perfusion of blood cell suspensions.
Figure1 A shows a representative set of pressure-flow curves obtained during perfusion of an isolated gracilis muscle. Because the zero-flow pressure in this muscle preparation did not change significantly (Table1) during perfusion with each suspension, it was subtracted from the measured arterial pressures. Such an approach permits a direct comparison of shifts in the pressure-flow curves due only to changing blood cell composition in the perfusate. Small variations in the zero-flow pressure are not due to varying blood cell concentration but are induced primarily by variations in the central circulatory hemodynamics of each individual animal (30). Finally, because the venous pressure was held at zero, this calculated pressure represents the driving pressure across the isolated muscle.
At a constant input flow rate, several additional pressure drops were defined to describe the relative effects of erythrocytes and leukocytes on the pressure-flow relationships. In these experiments (see Fig.1 A), an ES with 39% hematocrit induced a significant increase in arterial pressure above that measured during perfusion with the CFS. This additional pressure drop due to erythrocytes (ΔPE) reflects the difference in the apparent viscosity of these two suspensions. The large additional pressure drop induced by leukocytes in the presence of erythrocytes (ΔPLE) is comparable to previously observed values (29).
In contrast, when an isolated suspension of leukocytes containing the same cell concentration was perfused, only a slight additional pressure drop (ΔPL) was observed at each flow rate. Furthermore, the sum of the additional pressure drops induced by erythrocytes and leukocytes alone underestimates the additional pressure drop measured during perfusion with LES (Fig.1 B).
The relative additional pressure drop (ΔPREL), defined as the dimensionless ratio of ΔPL per leukocyte to ΔPE per erythrocyte, is shown in Fig. 2. This quantity serves to demonstrate that the relative effect of leukocytes compared with erythrocytes on the pressure-flow relationship was significantly different when LS was perfused in contrast to LES. The effect on organ hemodynamic resistance of each leukocyte compared with that of an erythrocyte when LS is perfused is significantly less than that obtained when LES is perfused. Furthermore, the large relative additional pressure drop induced by leukocytes seemed to depend on the presence of erythrocytes. In contrast, the relative additional pressure drop did not depend strongly on leukocyte concentration at constant hematocrit in the range of 3.6–6.2 × 106 leukocytes/ml (Fig.3).
Perfusion of microsphere suspensions.
Pressure-flow relationships during perfusion of the isolated gracilis muscle with microsphere suspensions were qualitatively similar to those measured during perfusion with blood cell suspensions. For example, Fig. 4 Ashows a representative set of pressure-flow curves obtained during perfusion of an isolated gracilis muscle. Pressure and flow axes are defined in the same manner as those described for blood cell perfusion.
At a constant input flow rate, additional pressure drops were measured to describe the relative effects of erythrocytes and microspheres on the pressure-flow relationships. In this experiment (Fig.4 A), an ES with 35% hematocrit induced a significant increase (ΔPE) in arterial pressure above that measured during perfusion with CFS. An MS containing 7.0 × 106 microspheres/ml caused an insignificant additional pressure drop (ΔPM). However, perfusion of the isolated gracilis muscle with an MES resulted in a large additional pressure drop (ΔPME) as a result of the presence of microspheres. This result was similar to that observed during perfusion with LES. Furthermore, the sum of the additional pressure drops induced by erythrocytes and microspheres alone (ΔPE + ΔPM) underestimated the pressure at a given flow rate as measured during perfusion with MES (Fig. 4 B).
In a manner similar to the analysis of blood cell perfusion shown above (Fig. 2), ΔPREL was computed as the dimensionless ratio of ΔPMper microsphere to ΔPE per erythrocyte. This computation enabled comparison of the relative effects of individual microspheres with those of erythrocytes on the pressure-flow relationships. At constant microsphere concentration, the magnitude of the additional pressure drop induced by microspheres was increased as the hematocrit was raised (Fig.5).
These results provide new insight into the impact of circulating leukocytes on organ perfusion in vivo. A shift of the perfusion pressure at constant flow indicates a change in blood flow resistance that may be the result of a change in plasma viscosity (31) or an altered apparent viscosity due to a local change in hematocrit (32). In line with previous observations (32), perfusion of a pure erythrocyte suspension induces an increased arterial pressure compared with the pressure measured during perfusion of a cell-free plasma substitute at the same arterial flow rate (Fig.1 A). The addition of leukocytes (<1% of blood cells by volume) to the erythrocyte suspension leads to a surprisingly large increase in perfusion pressure at each flow rate (29). However, when isolated leukocytes at the same low cell concentration are perfused without erythrocytes present, only a small shift in the pressure-flow curve is observed compared with that measured during perfusion with plasma substitute alone. Because the sum of the additional pressure drops due to isolated leukocytes and pure erythrocytes underestimates the additional pressure drop induced by a suspension containing both leukocytes and erythrocytes (Figs.1 B and 2), the hemodynamic resistance is clearly not determined solely by the concentration of blood cells in the perfusion medium.
Because the relative additional pressure drop due to isolated leukocytes is small compared with that induced by a mixed leukocyte-erythrocyte suspension, the net effect of leukocytes on hemodynamic resistance may be exhibited only in the presence of erythrocytes. Only a transient increase in flow resistance has been observed after a leukocyte bolus injection in the rat hindlimb (4), rat kidney (6), hamster cremaster muscle (11), and, to a lesser extent, rat lung (7, 36). In the rat hindlimb, the effects of leukocytes were exerted at or near the capillary networks, but only when an isolated leukocyte bolus was infused (4). Because of short capillary lengths in the lung, erythrocytes may collide repeatedly with the endothelial walls and may assume less frequently a preferred orientation, which can be disturbed by leukocytes. However, only transient effects induced by a leukocyte-erythrocyte bolus have been recorded (36). Furthermore, injection of a leukocyte-erythrocyte bolus into the hamster cremaster muscle induced a small transient increase in microvascular resistance, but a significantly larger increase was achieved when successive boluses containing leukocytes were infused (10). This indicates that not only are erythrocytes necessary for leukocytes to exert an effect on microvascular resistance but also that transient changes in microvascular resistance may differ from steady-state values.
To further illustrate the role of erythrocytes in a leukocyte-mediated effect on organ perfusion, an estimate of whole organ perfusion resistance was calculated as the reciprocal slope of the pressure-flow curve during perfusion with each suspension. The resistance during perfusion with the leukocyte-containing suspensions was normalized by the resistance due to the erythrocyte suspension at constant arterial flow rates (Fig. 6). This relative resistance ratio demonstrates that the resistance attributable to the presence of leukocytes when erythrocytes were also included in the perfusion medium was significantly higher than resistance induced by either isolated leukocytes or pure erythrocytes. Furthermore, the resistance during perfusion with isolated leukocytes was significantly lower than that induced by a pure erythrocyte suspension. These results parallel those demonstrated by comparison of the relative additional pressure drop attributable to leukocytes in the presence or absence of erythrocytes (Fig. 2).
A number of mechanisms involving hydrodynamic interactions between leukocytes and erythrocytes may explain the additional pressure drop attributable to circulating leukocytes. Because leukocytes are larger and less deformable than erythrocytes, they are carried through capillaries at a slower velocity than erythrocytes (21). As a result, the erythrocytes trapped behind a leukocyte must move with a slower velocity than if no leukocytes were present, and capillary hematocrit is increased. At the exit from the capillaries, erythrocytes squeeze past the leukocyte, displacing the leukocyte radially toward the vessel wall in the process (27). This margination serves to promote leukocyte adhesive interactions with the postcapillary endothelium and raises the resistance in venules (17). However, perfusion of microspheres instead of leukocytes in the gracilis muscle also produced a large shift in the pressure-flow curve in the presence of erythrocytes (Fig.4 A), suggesting that adhesion to the postcapillary venular wall did not play a significant role in the current studies.
The formation of erythrocyte “trains” behind more slowly moving leukocytes in capillaries may also serve as an important mechanism by which leukocyte-erythrocyte interactions enhance hemodynamic resistance. Erythrocytes on the upstream side of a leukocyte are crowded together and are displaced to an off-center radial position, whereas a cell-free region develops downstream of the leukocyte (27,33). Changes in the flow characteristics of erythrocytes may have a direct effect on the hemodynamic resistance and local blood rheology (8, 9, 16, 20). At low shear rates such as those present in the microcirculation, flow resistance is minimized when erythrocytes are aligned near the centerline of the vessel (2). As the axial velocity of erythrocytes is reduced by the more slowly moving leukocyte, the width of the plasma lubrication layer decreases, and the relative apparent viscosity of the blood increases (1). An effect of leukocytes on flow resistance can be demonstrated in an in vitro model with relatively small cell concentrations (17.7% hematocrit and 106 leukocytes/ml) (33), indicating that a mechanism involving changes in erythrocyte flow characteristics may be sensitive to small changes in local tube hematocrit, even when only few leukocytes are present. Because the relative additional pressure drop measured in the isolated gracilis muscle preparation appears to be relatively independent of the exact leukocyte concentration within the physiological range (Fig. 3), only comparatively few leukocytes in microvessels may be needed to disturb the flow of erythrocytes.
Earlier in vitro studies have concluded that leukocytes may serve to substantially increase flow resistance via their effects on erythrocyte train formation (12). Recently, an “in vivo viscosity law” has been proposed to allow the prediction of microvascular resistance in good agreement with calculated resistance values based on in vivo pressure and flow measurements (24). The contribution of leukocytes to blood apparent viscosity in the microcirculation was not explicitly considered in the mathematical derivation of relative apparent viscosity, but the effects of leukocytes may have been included in the estimate of empirical parameters. In another study under in vivo conditions, erythrocyte velocity was equal to leukocyte velocity in a capillary network where both cell types were present (10). Thus measurements in vessels without leukocytes may be required to accurately elucidate the specific effects of leukocytes on microvascular resistance.
Direct in vivo observation of the cat retinal microcirculation provided evidence that erythrocyte velocity was decreased in the presence of leukocytes, especially in “tortuous” capillaries (3). Increased tortuosity may reduce erythrocyte velocity or increase axial erythrocyte displacement independently of leukocyte effects. Because skeletal muscle capillaries are relatively straight and parallel compared with those in the retina, the effects of leukocytes on erythrocyte velocity may be enhanced in skeletal muscle. The effect of leukocytes on perfusion resistance may be organ dependent.
To investigate whether a mechanism for the leukocyte-induced increase in perfusion resistance depends on the physical presence of leukocytes per se, the leukocytes were replaced in some experiments by rigid polystyrene microspheres. Perfusion of a suspension containing only microspheres had essentially no effect on the pressure-flow relationship (Fig. 4 A), whereas the isolated leukocyte suspension slightly increased the pressure at each flow rate. In contrast to microspheres, leukocyte adhesive mechanisms may enhance hemodynamic resistance compared with plasma in the absence of erythrocytes. However, in a manner similar to that of the leukocytes, the microspheres cause a marked increase in perfusion resistance only in the presence of erythrocytes. In line with the mechanism proposed for mixtures of leukocytes and erythrocytes, the rigid microspheres may also cause a decrease in erythrocyte velocity and a radial displacement to an off-axis position in capillaries. In addition, perfusion with mixed microsphere-erythrocyte suspensions with different erythrocyte concentrations demonstrated that the magnitude of the relative additional pressure drop induced by microspheres depended on the hematocrit (Fig. 5).
A direct comparison of relative resistance was also calculated for microsphere and erythrocyte suspensions (Fig.7). The relative resistance induced by a pure microsphere suspension was significantly lower than when erythrocytes were present in the perfusion medium. Furthermore, the relative resistance increase induced by microspheres in the presence of erythrocytes depended on hematocrit. Thus mechanisms that enhance organ perfusion resistance appear to depend on blood cell particle flow dynamics and not only on leukocyte physiological properties per se.
In summary, the present study demonstrates that leukocytes contribute to an increased organ hemodynamic resistance in the skeletal muscle circulation more than can be attributed to leukocytes alone. A mechanism is proposed in which leukocytes disrupt the normal flow patterns of erythrocytes, especially in the capillaries and postcapillary venules. This is primarily a rheological phenomenon and depends on the physical sizes and properties of leukocytes and erythrocytes. Thus leukocytes make an important contribution to organ hemodynamics, even though they constitute <1% of the blood cells in the circulation. By this new mechanism, leukocytes may serve to significantly increase blood flow resistance in both physiological and pathophysiological conditions.
This study was supported by National Heart, Lung, and Blood Institute Grant HL-10881 and National Science Foundation Grant IBN-9512778.
Address for reprint requests: B. P. Helmke, Institute for Medicine and Engineering, University of Pennsylvania, 3508 Market St., Suite 380, Philadelphia, PA 19104-3357.
- Copyright © 1997 the American Physiological Society