| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Bioengineering Program, The Pennsylvania State University, University Park, Pennsylvania 16802
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The effect of
reduced red blood cell (RBC) deformability on microvessel recruitment
attendant to a reduction in tissue
PO2 was studied in rat cremaster
muscle using indicator-dilution techniques. Transit times (TT) of
fluorescently labeled RBCs
(TTRBC) and plasma (TTPl) between functionally
paired arterioles and venules were obtained from their dispersion
throughout the microvascular network. Changes in
PO2 were effected by superfusing the
tissue with Ringer solution deoxygenated to different levels.
Arteriolar blood flow (
) was measured with the
two-slit technique, and the vascular volume (V) occupied by RBCs and
plasma was computed from the product of × TT
during bolus infusions of rat and less deformable human RBCs to obtain
VRBC and fluorescently labeled albumin to obtain VPl.
Measurements of TTRBC and
TTPl permitted computation of an
average flow-weighted tissue (microvascular) hematocrit
(HM) relative to systemic values
(HS). During
infusions of autologous rat RBCs, and total V
increased threefold in response to hypoxia, whereas normalized RBC TT
(TTRBC/TTPl)
and normalized tissue hematocrit
(HM/HS)
did not show a significant trend, indicating an increase in the number
of pathways through which the RBCs can traverse the network because of
spatial recruitment of capillaries. In contrast, during infusions of
human RBCs,
TTRBC/TTPl and
HM/HS
decreased significantly in response to hypoxia. Although exhibited an increase similar to that during
rat RBC infusions, VRBC exhibited
a smaller increase compared with
VPl, suggesting that reduced RBC
deformability leads to a redistribution of RBCs through larger-diameter
pathways within the network and exclusion of these RBCs from pathways
normally recruited during hypoxia. Hence, reduced RBC deformability may
adversely affect capillary recruitment and physiological mechanisms
that ensure adequate delivery of oxygen to tissue.
vascular volume; erythrocyte deformability; tissue hematocrit; microvascular perfusion; hypoxia; oxygen transport
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
IT HAS BEEN FIRMLY established that diminished oxygen delivery to tissue in response to hypoxia is countered by a combination of increased regional blood flow and increased functional capillary density in the microcirculation (11, 12, 17). It has been demonstrated that the magnitude of these compensatory responses depends on network topography (18, 26), initial vascular tone (9), and, indirectly, red blood cell (RBC) mechanical properties (31). In an attempt to elucidate the relative roles of hyperemia and capillary recruitment, it has been demonstrated that significant capillary reserves may be recruited during reactive hyperemia (34), exercise (22, 7), anoxia (18), and neural stimulation (6). Many of these studies used direct observations of alterations in capillary density and flow in exteriorized tissues that attempted to establish a quantitative assessment of capillary recruitment and flow. For example, Prewitt and Johnson (24) and Klitzman et al. (9) made direct measurements of alterations in capillary density in response to diminished levels of tissue PO2 by superfusing the tissue with Ringer solution equilibrated to various levels of PO2. Although such studies offer valuable insight into the extent of capillary recruitment by computing the numbers of capillaries perfused per unit volume of tissue, they are often limited by the inability to account for all capillaries served by a specific pair of arterioles and venules that serve a discrete region of tissue. Furthermore, a judgment of whether a capillary is perfused must be made irrespective of the actual flux of RBCs carried within each capillary, and thus estimates of capillary density may represent a best-case situation.
Studies of the rheological behavior of RBCs in the capillary network
clearly demonstrated that capillary RBC flux and velocity are strongly
dependent on the ability of RBCs to deform on entry into the
capillaries. It was clearly demonstrated that reductions in RBC
deformability may adversely affect capillary perfusion (4) and that
many diseases manifest reductions in RBC deformability. For example,
elevated internal fluid viscosity or abnormal membrane stiffness was
found in diabetes mellitus (28), 
- and 
-thalassemia (29), and
sickle cell disease (27). Of the many determinants of capillary
perfusion, the size of the undeformed RBC relative to the capillary
diameter may play the greatest role in affecting capillary perfusion.
For example, studies of the passage of RBCs through capillary-sized
pores of polycarbonate sieves (25) reveal that the ratio () of the
resistance to flow of suspensions of RBCs
(RRBC) to that
of their suspending medium
(RPl) may
increase 30-40 times as the ratio of pore to cell diameter is
reduced from 1 to 0.1. Furthermore, after entry into a capillary, the
ability of RBCs to deform may play an equally important role as RBCs
negotiate irregularities in the capillary lumen, as manifested by
encroachment of endothelial cell nuclei on the capillary lumen (30).
It was also demonstrated that the microvascular network may passively
compensate for increased RBC stiffness by shunting RBCs within the
capillary network through pathways of lesser resistance (14). As shown
in that study, using RBCs hardened to various degrees with
glutaraldehyde, the transit time (TT) of hardened RBCs in cremaster
muscle remained unaffected as was increased from a norm of 2.6 (using polycarbonate filters with a mean pore diameter of 4.7 µm) to
10.0. However, as was increased from 10 to 20, the multitude of
pathways within the cremaster muscle were overwhelmed, and TT across
the capillary network rose threefold. Although it is now clearly
recognized that capillary perfusion may be attenuated because of
diminished RBC deformability, it is generally unknown to what extent
these abnormalities may affect microvascular function in light of the
ability to recruit capillaries in response to hypoxia.
To fill this void, the present study uses a combination of techniques
to permit calculation of the total vascular volume
(VTot) within the
microvasculature proper served by a well-defined pair of terminal
arteriole and collecting venule. Using the two-slit photometric
technique (33) to measure RBC velocity
(VRBC) in the
arteriole and venule in the cremaster muscle (rat), we calculated the
volumetric flow () of blood delivered to discrete
groups of capillaries. Concomitantly, techniques of indicator dilution (14) were applied to estimate RBC and plasma mean TT across the
discrete units of capillaries served by these microvessels. The
vascular volume of all true capillary vessels between paired arteriole
and venule was then calculated from the product × TT and examined in light of reductions in tissue
PO2 and RBC deformability. To model
the effects of reduced RBC deformability, these measurements were made
during infusion of both autologous RBCs from donor animals and larger
RBCs from human subjects.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The procedures for acquisition and analysis of indicator TT are
summarized in Fig. 1, and detailed methods
are as follows.
|
Animal preparation. Male Sprague-Dawley rats 33-38 days old and weighing between 70 and 100 g were anesthetized with pentobarbital sodium (35 mg/kg ip), tracheostomized, and allowed to breathe spontaneously. The right internal jugular vein was cannulated with polyethylene (PE-50) tubing for administration of supplemental doses of anesthetic at a rate of 7 mg/kg every 30 min, or as required, to maintain a surgical plane of anesthesia. The left carotid artery was cannulated with PE-50 tubing and connected to a strain gauge-type pressure transducer to monitor arterial blood pressure. The left femoral artery was cannulated with PE-10 tubing for injection of fluorescent indicator. Systemic hematocrit (HS) and white blood cell (WBC) and RBC counts were measured by withdrawing 0.1-0.2 ml of blood through the carotid catheter. For the 20 animals studied, WBC count averaged 8,659 ± 2,672 (SD) cells/µl, and HS averaged 0.368 ± 0.016 (SD). The animal was placed on a stage heated to 37°C by circulating water. The right cremaster muscle was surgically exposed, spread over a glass pedestal, and maintained at 35°C by superfusion with HEPES-buffered Ringer solution at a rate of 25 ml/min. The arterial blood pressure was maintained between 100 and 130 mmHg, and the initial blood pressure averaged 109.4 ± 17.84 (SD) mmHg.
Fluorescent RBC and Plasma Preparation
Erythrocytes from rat and human subjects were fluorescently labeled to facilitate the application of indicator-dilution methods for the determination of mean TT. Male Sprague-Dawley rats, 85-105 days old and weighing between 300 and 450 g, were anesthetized with pentobarbital sodium (35 mg/kg ip), and the internal jugular vein and the carotid artery were catheterized with polyethylene (PE-90) tubing. Heparin was administered intravenously (500 U · ml
1 · kg1),
and after 5 min the animal was exsanguinated through the carotid catheter. To obtain human RBCs, healthy human donors from our laboratory (25-35 yr) were selected and 10 ml of whole blood was withdrawn by venipuncture into heparinized tubes.
Methods for preparation of fluorescently labeled RBCs (FRBCs) were described previously (14). In brief, whole blood was washed by repeated centrifugation (850 g) in Tris-buffered Ringer solution (pH = 7.4). Fluorescent dye solution was prepared by solubilizing 20 mg of FITC in 0.5 ml of DMSO and adding this solution to 50 ml of Tris-buffered Ringer solution containing 0.5% bovine albumin. The packed RBCs were added to the dye solution and incubated for 1 h at room temperature. The FRBCs were washed twice with Tris-buffered Ringer solution and once with Tris-buffered Ringer solution containing 1% bovine albumin (TRA, pH = 7.4). The cells were filtered through a 12-µm polycarbonate filter to remove clumps of cells and then resuspended in TRA. Each suspension of RBCs was adjusted to a hematocrit that matched the HS of the recipient animal and warmed to 37°C before infusion.
To facilitate the determination of plasma TT, a 1% stock solution of tetramethylrhodamine isothiocyanate-albumin in saline was diluted with an equal volume of heparinized saline immediately before infusion.
Experimental Protocols
The cremaster muscle was allowed to stabilize for 30 min before data acquisition. During this period, the superfusate was aerated with room air and its PO2 was maintained at high values (typically 110 mmHg). A functionally paired arteriole and venule were identified by tracing the flow pathways between them and verifying that conservation of flow was maintained between the two vessels. The selected arteriole-venule pairs consisted of a terminal arteriole and collecting venule that bordered a recognizable group of true capillaries as illustrated in Fig. 2. As shown, the passage of a fluorescent bolus clearly demarcated a group of capillaries within a well-defined region of tissue. To minimize the error in TT calculations (14), only vessel pairs with arteriolar and venular blood flow within 20% of each other were used in this study. The arteriole was viewed under high magnification using a Leitz UM20 objective with a magnification of ×13 and numerical aperture (NA) 0.22. Arteriolar and venular VRBC and arteriolar diameters (DA) were measured, and the magnification was reduced to acquire a larger field of view using a Leitz UM10 objective with magnification of ×6.5 and NA 0.14. To examine the role of vascular tone, a separate set of experiments following the same protocol were performed using networks vasodilated with 104 M
papaverine added to the superfusate.
|
Measurement of PO2
The PO2 on the surface of the cremaster muscle was measured using a membrane-covered 18-gauge platinum electrode (model 760S, Diamond General), calibrated over the range of 0-156 mmHg, by bubbling gases containing 0, 5, 10, and 20% oxygen through Tris-buffered Ringer solution placed in a constant-temperature (35°C) bath. The platinum electrode was held in a micromanipulator, and its tip was placed within 1,000-2,000 µm of the arteriole-venule pair of interest and away from any large arteriole or venule. The electrode tip was positioned 50-100 µm above the tissue and maintained at the same position throughout the experiment. The proximity of the electrode tip to the tissue ensured that the PO2 measurements were representative of PO2 in the tissue. The choice of this relatively large electrode was dictated by the goal of rapidly obtaining a representative index of tissue oxygenation without disturbing the microcirculation. A reference Ag-AgCl electrode was inserted into the abdominal cavity through a small incision on the lower right abdominal wall above the right femoral artery. Tissue PO2 was reduced to desired levels by varying the oxygen content of the superfusion solution (to ~70, 40, and 10% of initial value) by bubbling a mixture of 95% N2-5% CO2 through the superfusate reservoir. The tissue was allowed to stabilize for several minutes as PO2 was incrementally reduced.Prior measurements of tissue PO2 (5,
9, 24) demonstrated that changes in tissue
PO2 faithfully follow changes in
superfusate PO2. Moreover, it has
been shown previously that with sufficiently high superfusion rates (25 ml/min) and diminished levels of superfusate
PO2, all oxygen may be washed out of
the tissue (16). As shown in that study with measurements of
perivascular PO2 using gold-filled micropipettes, PO2 could be
effectively reduced to values near zero when the superfusate was
equilibrated with a mixture of 95%
N2-5%
CO2. To verify that changes in
superfusate PO2 reflected
commensurate reductions in tissue
PO2, perivascular tissue
PO2 was measured in separate
experiments (n = 3) using gold-filled
micropipettes (Diamond General) with a tip diameter of ~5 µm and
compared with simultaneous measurements of superfusate
PO2 using the platinum electrode. The micropipettes were placed within the muscle interstitial tissue space
in a periarteriolar position that ranged from 10 to 50 µm from the
feeding arteriole. The results (Fig. 3)
demonstrate that although tissue PO2
may be ~35% below that of the superfusate at high
PO2 (presumably due to tissue oxygen
consumption and convection of oxygen away by blood flow), the disparity
between the two measurements is eliminated as superfusate
PO2 falls below 10 mmHg. Thus the
superfusate values of PO2 measured on
the tissue surface provide a realistic quantitative index of the extent
of tissue oxygenation.
|
Measurement of Volumetric Flow
The centerline VRBC of blood in arterioles and venules was measured using the two-slit photometric technique (33). The mean velocity of cells plus plasma within arterioles (VA) and venules (VV) was calculated as the centerline V divided by 1.6. DA and diameter of venules (DV) were measured using the video image shearing technique. The volumetric flow of blood within the arterioles (A) and
venules (V)
was calculated as the product of mean V and cross-sectional area,
D2/4.
Indicator-Dilution Techniques
TT of fluorescently labeled plasma (TTPl) and RBCs (TTRBC) were calculated using indicator-dilution techniques as described previously (19). The hematocrit and cell count of the FRBCs were matched to those of the recipient animal using a Coulter counter. Bolus injections of 0.05 and 0.025 ml of RBCs and plasma, respectively, were made through the femoral catheter contralateral to the cremaster muscle. Three successive boluses of fluorescently labeled plasma followed by three boluses of FRBCs maintained at 37°C were injected, and the dispersion of the indicator through the selected arteriole-venule pair was recorded on a videotape (Fig. 1). The superfusate PO2 was reduced, and after allowing sufficient time for the tissue to stabilize at the new PO2, we repeated the bolus injections of RBCs and plasma. The intensity-time curves of each bolus passing through the arteriole-venule pair were obtained by videodensitometry using an IPM video photoanalyzer (model 204, Instrumentation for Physiology and Medicine, San Diego, CA). This device provides an analog signal proportional to video scene intensity averaged over a photometric window, as illustrated in Fig. 1. Each of the two photometric windows was placed over the arteriole and venule, respectively, to acquire a signal representative of the average intensity across the lumen, as shown in Fig. 2. Each intensity-time curve was digitized to calculate TT by the cross- correlation technique (19).Computation of Vascular Volume and Tissue Hematocrit
The volumes of the RBC (VRBC) and the plasma (VPl) compartments served by the selected arteriole-venule pair were calculated from the product of mean TT of the indicator and (20), namely
|
|
|
Estimates of Capillary Density
To facilitate a comparison of the present measurements of vascular volume (V) with direct measurements of capillary density found in the literature, the relationship between V and capillary density per unit volume (
) was expressed in terms of the area of the tissue served by
the feeding arteriole-venule pair
(At), the
thickness of the tissue (
), and the volume of an individual capillary (VC), namely
|
with hypoxia, changes in vascular V with
PO2 were calculated using the slope
of the linear regression curve of absolute
VTot plotted as a function of
PO2 and the effective variation of
determined from the slope. The maximum area of tissue at the lowest
PO2, perfused by all capillaries
between the selected arteriole-venule pair, was measured in square
micrometers, using image-processing software (Image Tool, University of
Texas, Southwestern Medical Center). VC was estimated using previously
published values of mean capillary length and diameter of 615 and 6 µm, respectively (32). Cremaster muscle thickness was obtained from
previously published data on rat cremaster muscle (1) as 175 ± 15 µm.
Statistics
Linear regression was used to assess the trend of a given variable. The parameters of linear regression for all variables are presented in Table 1. The slope of the linear regression was considered significant for P < 0.05 (t-test).
|
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
TT and hemodynamic measurements were made during reductions in
PO2 for rat FRBCs and human FRBCs in
separate cremaster networks with normal vascular tone and vasodilated
(papaverine) conditions. The initial mean arterial blood pressure
during rat RBC experiments was 104 ± 16 and 122 ± 15 mmHg for
normal tone and vasodilated preparations, respectively, and that during
experiments with human RBCs was 100 ± 12 and 111 ± 10 mmHg,
respectively (Table 2). Initial
HS for experiments with rat RBCs
and human RBCs was 0.37 ± 0.008 and 0.35 ± 0.02, respectively,
in normal tone networks and 0.36 ± 0.01 and 0.36 ± 0, respectively, in dilated networks (Table 2).
|
Control Values
To minimize the effects of heterogeneity among preparations, all variables were normalized with respect to their initial values. To permit comparison with results from similar studies, a list of initial values of all variables is reported in Table 2, in which initial values for experiments using rat RBCs in networks with normal tone and dilated networks and human RBCs in normal tone networks are presented. The diameters and velocities presented here were obtained from fourth-order arterioles and venules. The initial diameters observed in the literature for similar-order vessels are varied and reflect the user-specific definition of the order of observed vessels. Studies by McKay and Lipowsky (19) show the average diameter for fourth-order vessels to be 23 ± 6 (arterioles) and 30 ± 18 (venules) µm, whereas Lipowsky et al. (14) present an average diameter of 26 ± 7 (arterioles) and 34 ± 7 (venules) µm for fourth-order vessels. TTRBC and TTPl obtained for normal tone rat vessels in the present study were lower than those observed by McKay and Lipowsky (19) for fifth-order vessels and lower than those observed for fourth-order arteriole-venule pairs (14). VTot vascular volume calculated in the present study was higher than those calculated from studies by Klitzman et al. (9) and le Noble et al. (13) and lower than that obtained from Honig et al. (7). These differences could be attributed to the different species and animals used: hamster cremaster (9), Wistar-Kyoto rat cremaster (13), and dog gracilis (7). The values of HM/HS for normal tone vessels were almost twice those measured by House and Lipowsky (8) using direct cell counting and are in good agreement with prior determinations using TT measurements in fourth-order vessels (14).Tissue PO2
Tissue PO2 was considered the independent variable, and all results obtained in this study were expressed as a function of PO2. At control conditions, PO2 was 105 ± 20 and 109 ± 15 mmHg for normal tone networks using rat RBCs and human RBCs, respectively (Table 2). For vasodilated networks, the values were slightly higher at 120 ± 9 (rat RBCs) and 122 ± 12 (human RBCs) mmHg. Prior studies using a similar setup for measuring PO2 recorded PO2 values of 85 ± 5.8 (24) and 133.2 ± 5.8 (9) mmHg when the superfusate was equilibrated with room air, reflecting the heterogeneity in these measurements. After complete deoxygenation, PO2 was 15 ± 11 (rat RBC experiments) and 14 ± 5 (human RBC experiments) mmHg for normal tone networks and 9 ± 5 (rat RBC) and 16 ± 2 (human RBC) mmHg for vasodilated networks. These values were comparable to those observed in earlier studies: 12 ± 2.3 (9) and 24 ± 1.2 (24) mmHg.To qualitatively illustrate the effect of PO2 reductions on network perfusion, video scenes during dispersion of FRBCs at high PO2 (130 mmHg) and low PO2 (35 mmHg) are presented in Fig. 2. The area of tissue supplied by all capillaries between arteriolar and venular measurement sites is clearly increased as tissue PO2 decreases.
Rat FRBCs in Normal Tone Networks
Hemodynamic variations.
To characterize the hemodynamic changes within the selected
arteriole-venule pairs in response to tissue hypoxia, the observed changes in velocity and diameter and the calculated changes in for both arterioles and venules were
plotted as a function of the absolute values of measured
PO2 (Fig.
4). Mean VA (Fig.
4A) increased significantly to 190%
of the control value (P < 0.05),
whereas mean VV
(Fig. 4B) increased to 220% of its control value (P = 0.11) in response
to the graded reduction in PO2. This
increase in VA is
50% less than that observed by Prewitt and Johnson (24) for arterioles
with diameters in the range of 10 to 30 µm. This lesser increase may
be attributed to the larger initial velocity: 13.08 ± 8.09 mm/s in
the present study compared with 0.69 ± 0.3 mm/s (24). Mean
DA (Fig.
4C) increased by 60%
(P < 0.0005), whereas mean
DV (Fig.
4D) increased by 26%
(P < 0.005). Comparable increases in
DA were recorded
during prior studies using superfused hamster cremaster muscle (9). Calculated values of volumetric flow increased 3.5 times for both arterioles (Fig. 4E) and venules
(Fig. 4F) and were significant at
P < 0.005 and
P < 0.05, respectively.
|
TT measurements.
Mean TTRBC and
TTPl are shown in Fig.
5 after reductions in
PO2. Rat
TTRBC (Fig.
5A) decreased 10%
(P > 0.6) from its control value,
whereas rat TTPl (Fig.
5B) increased 5%
(P > 0.9) from its control value, in
response to hypoxia. Because of the variability among preparations, rat
TTRBC and rat
TTPl did not follow a significant
trend. To reduce the ambiguity introduced by this heterogeneity, the
ratio of paired rat RBC and plasma TT measurements
(TTRBC/TTPl)
was calculated and plotted as a function of
PO2.
TTRBC/TTPl
for rat RBCs (Fig. 5C) increased by 25% and also did not show a significant trend
(P > 0.7). This increase, however,
suggests that RBC residence time within the capillary network increased
more than plasma residence time, as would be expected under hypoxic
conditions.
|
Volume and hematocrit measurements.
Changes in vascular volume and microvascular hematocrit are shown in
Fig. 6 for RBCs, plasma, and their sum. As
expected, VRBC (Fig.
6A),
VPl (Fig.
6B), and, hence,
VTot (Fig.
6C) increased about threefold above
their control values, suggesting an increase in the number of open
pathways with reduction in PO2. All
three increases were significant at P < 0.01 (VRBC),
P < 0.05 (VPl), and P < 0.01 (VTot). The average
HM (Fig.
6D) increased 10%, but the trend
was not significant (P > 0.75).
|
Human RBCs in Normal Tone Network Preparations
The results obtained for human RBC infusions in normal tone networks are plotted in Figs. 7 and 8. For the sake of brevity, only and normalized TT are shown in Fig.
7. In response to hypoxia,
A (Fig.
7A) increased 2.25 times the initial
value (P < 0.005).
Normalized TTRBC (Fig.
7B) for human RBCs decreased by
>25% (P < 0.01), indicating that
the residence time of human RBCs in rat vascular network increases at a
lower rate than that of plasma. Changes in vascular volume and
microvascular hematocrit with hypoxia are presented in Fig.
8. VPl
(Fig. 8A) increased >375%
(P < 0.005), whereas
VRBC (Fig.
8B) increased only 260% (P < 0.005), indicating
that the increase in number of capillaries traversed by human RBCs was
less than that traversed by plasma. VTot (Fig.
8C) increased 340%
(P < 0.001), and this increase was similar to that observed for rat RBCs in normal tone networks. The
trend observed by rat
TTRBC/TTPl
(Fig. 7B) was echoed by human HM/HS
(Fig. 8D), which decreased
significantly (P < 0.05) more than
20%, indicating that the larger human RBCs were excluded from the rat
vascular network.
|
|
Comparison of TTRBC/TTPl and HM/HS for Rat and Human RBC
The absolute (nonnormalized) values of TTRBC/TTPl (Fig. 9A) and HM/HS (Fig. 9B) are presented as a function of absolute PO2. Although TTRBC/TTPl for rat RBCs decreased from 0.75 ± 0.4 to 0.68 ± 0.34 (P = 0.27) as the PO2 decreased from 105 ± 20 to 15 ± 11 mmHg, TTRBC/TTPl for human RBCs decreased from 1.04 ± 0.4 to 0.72 ± 0.24 (P = 0.29) in response to a reduction in PO2 from 108 ± 16 to 14 ± 6 mmHg. For the same changes in PO2, HM/HS for rat RBCs increased from 0.74 ± 0.26 to 0.83 ± 0.17 (P = 0.37), whereas HM/HS for human RBCs decreased from 0.99 ± 0.25 to 0.79 ± 0.2 (P = 0.34).
|
Vasodilated Network Preparations
Changes in normalized TTRBC and HM for vasodilated networks subjected to infusions of rat and human RBCs are presented in Fig. 10. For rat RBCs, both TT (Fig. 10A) and hematocrit (Fig. 10B) did not show any significant trend with hypoxia (P > 0.75 and P > 0.8, respectively). For human RBCs, both TT (Fig. 10C) and hematocrit (Fig. 10D) recorded an increase, but the increases were not significant (P > 0.1 and P > 0.2, respectively). Volumetric flow and vascular volume also did not change significantly with reductions in PO2 in either rat or human RBCs (results not shown).
|
Capillary Density
Calculated variations in capillary density for rat RBCs in normal tone and vasodilated networks and human RBCs in vasodilated networks are presented in Table 3. Capillary density per unit area for rat RBCs in normal tone networks increased more than threefold from 161 capillaries/mm2 at 125 mmHg to 522 capillaries/mm2 at 10 mmHg. Similarly, capillary density per unit volume increased from 920 to 2,984 capillaries/mm3. For experiments with human RBCs in normal tone networks, capillary density per unit area increased from 109 to 345 capillaries/mm2 and capillary density per unit volume increased from 621 to 1,973 capillaries/mm3 for the same changes in PO2. In contrast, the capillary density of vasodilated networks increased only 40%, from 257 to 350 capillaries/mm2. To establish a frame of reference, total capillary volume and capillary density per unit area and per unit volume for two prior studies are also presented in Table 3. To perform a comparison, estimates of parameters measured here were derived from these earlier data. Prewitt and Johnson (24) counted the number of capillaries in a single direction perpendicular to the capillaries to obtain capillary density per millimeter. Using the method therein, capillary density per unit area was derived by taking the square of this value. However, Klitzman et al. (9) recorded capillary density per unit area, and hence, using single-capillary volume data calculated earlier (see METHODS), the total capillary volume was back-calculated. It is evident that these studies agree favorably with the current results with differences as noted in the DISCUSSION.
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The present study has aimed to elucidate the effects of changes in RBC
deformability on microvascular changes in capillary density caused by
tissue hypoxia in the cremaster muscle of the rat. Previous studies of
the relationship between capillary density and tissue
PO2 relied on direct counting of
individual capillaries to obtain measures of capillary density (9, 24). However, these measurements do not account for flow rates within individual pathways from arteriole to venule. In contrast, the present
study utilizes the techniques of indicator dilution, which offer an
indirect measure of capillary density that fully reflects the magnitude
of flow within all pathways perfused within the true capillary network.
As illustrated by Meier and Zeirler (20), the mean TT of indicator
through the network represents a distribution of TT through all
possible pathways, and hence alterations in RBC velocity and the number
and length of these pathways would be reflected in the measured TT.
Hence, the indicator-dilution method offers a flow-weighted measure of
capillary density that accounts for capillaries with slowly or rapidly
moving RBCs, as well as capillaries outside the immediate field of
view. However, for the indicator-dilution method to accurately reflect
the volume and length of the pathways traversed by the indicator, it is
necessary to maintain flow conservation between the arteriole-venule
pair under consideration (14). As stated in
METHODS, the reported values of
A and
V agreed within
20% of each other. For 20 arteriole-venule pairs
A/V
averaged 1.17 ± 0.36 (SD), which was not significantly different
from 1.0 (P < 0.05). According to
the error analysis presented previously (14), the TT measurements should be accurate to within 15% and the estimates of fractional hematocrit accurate to within 30%.
The present results clearly demonstrate that a continuous trend of flow-weighted capillary density with hypoxia can be established using the techniques of indicator dilution. Although these trends are consistent with the previous studies (9, 21, 24), there are some methodological differences. Klitzman et al. (9) observed a 150% increase in capillary density for a reduction in superfusate PO2 from 133 to 12 mmHg using the resting hamster cremaster muscle model (Table 3). Morff (21) measured the intercapillary distance (ICD) in rat cremaster muscle placed in a bath containing bicarbonate-buffered Krebs solution and observed nearly a twofold decrease in ICD. Converting ICD to its equivalent capillary density, the twofold increase equates to a 330% increase in capillary density for a PO2 reduction from 73.5 to 8.4 mmHg. Prewitt and Johnson (24) reported a fourfold decrease in capillary density (Table 3) for the rat cremaster muscle as the superfusate PO2 was increased from 38 to 97 mmHg. These latter two studies (21, 24) agree well with the present study of rat RBCs.
The role of capillary hematocrit as a determinant of oxygen transport to tissue has been a subject of considerable interest, particularly in light of the well-known disparity between tube and discharge hematocrit (3). The technique of indicator dilution offers a unique opportunity to delineate a functional tissue hematocrit that represents a flow-weighted measure of the volume fraction of RBCs resident within the capillary network. In contrast to direct measurements of capillary hematocrit by RBC counting within individual capillaries, the flow-weighted tissue hematocrit accounts for the presence of all pathways available for RBC and plasma flow, regardless of their velocities or the necessity of determining which capillaries carry off the majority of the network throughput. The present measurements of microvascular hematocrit, calculated from TTRBC and TTPl, suggest that average HM remains invariant with onset of hypoxia and capillary recruitment. A nonsignificant (P > 0.2) 12% increase in HM/HS was found in response to the papaverine-induced localized vasodilation (Table 1), and a small yet statistically nonsignificant 12% increase in HM/HS from 0.74 to 0.83 was found for networks with normal vascular tone after reductions in tissue PO2 (Fig. 9B). Although this latter increase in hematocrit was inversely proportional to tissue PO2, as obtained by direct cell counting (3, 10), the magnitude of the response was markedly different. Direct cell counting measures of capillary hematocrit in resting cremaster muscle at normal PO2 have revealed resting values of HM/HS ranging from 0.12 (10) to 0.48 (8). With reductions in tissue PO2, a significant increase in HM/HS (~65%) was observed as PO2 was decreased from 40 to 8 mmHg (10), with values of HM/HS increasing from 0.12 to 0.20, respectively. However, the present results are also consistent with the observed invariance in capillary hematocrit obtained by direct cell counting during postocclusive reactive hyperemia (8). In contrast, direct cell counting methods also showed a fourfold increase in capillary hematocrit in response to repeated muscular contraction and vasodilation with adenosine (10). This protocol also resulted in an increase in arteriolar flow that was twice the magnitude of that found here concomitant to hypoxia alone and also twice that found during postocclusive reactive hyperemia (8). Thus it appears reasonable to conclude that the functional hyperemia incurred with exercise produces a vasodilation in arterioles proximal to the sites affected by the application of papaverine or postocclusive reactive hyperemia (8). It is conceivable that a dramatic increase in capillary hematocrit may ensue as pressure gradients increase across the capillary network with the dilation of large arterioles, which facilitates the perfusion of newly recruited capillaries.
It should be noted that the smaller values of HM/HS obtained from direct measurements, compared with values derived from TT measurements, may be caused by discrepancies introduced by inadequate sampling of all possible pathways within the network. The present measurements using indicator-dilution methods agree closely with similar indicator-dilution measurements of HM/HS of 0.87 and 0.78 (14, 23), found in the resting state at high tissue PO2.
To establish a frame of reference for assessing the extent to which RBC
deformability affects capillary perfusion, many studies used the
filterability of RBC suspensions through the 5-µm pores of Nuclepore
filters. With this technique, the ability of the RBC to deform and
enter a capillary-sized tube may be characterized by the parameter ,
defined as the ratio of the resistance to flow through the filter pores
of a suspension of RBCs
(RRBC) to that
of the suspending medium
(RPl), i.e.,
RRBC/RPl.
The average value of for rat RBC has been shown to equal 2.61 ± 0.56 (SD) (14), whereas that for human RBC equals 4.3 ± 0.4 (27).
This difference in implies that the larger human cells (diameter = 7.82 µm, compared with 6.3 µm for rat RBCs) appear less deformable than autologous rat RBCs as they traverse the rat cremaster muscle microvasculature. In response to reductions in RBC deformability, Simchon et al. (31) observed entrapment of cells within the microvasculature and a concomitant reduction in flow through various tissues, thus suggesting that less deformable RBCs would encounter an
increased resistance within the capillaries. It has been shown, however, that the vascular network compensates for small increases in
by redistributing RBCs through pathways of least resistance as long
as is <10 (14). The present study demonstrates that the number of
such pathways through which the human RBCs can be redistributed and can
be recruited in response to hypoxia becomes diminished with reductions
in RBC deformability, as evidenced by the smaller increase in human
TTRBC compared with that of
TTPl (Fig.
7B). In addition, the similar
increase in capillary density of rat RBC- and human RBC-infused
networks suggests that the reduction in normalized TT
(TTRBC/TTPl)
is indicative of the exclusion of human cells from most of the normally
recruited pathways.
Because of the limited dispersion of human RBCs throughout the vascular
network, the microvascular hematocrit was found to decrease in response
to hypoxia (Fig. 8D). In contrast,
HM/HS for rat RBCs increased 12% from 0.74 to 0.83 in normal tone networks (Fig. 9B). However, the initial
value of
HM/HS
measured was 1.022 for tissues perfused with human RBCs and was greater
than that for rat RBCs (Fig. 9B).
Similarly, prior studies using indicator-dilution techniques (14) found
HM/HS
to rise from 0.87 to ~1.3 for glutaraldehyde-hardened RBCs as rose from 2.6 to 10. With further reductions in above a value of
10, HM/HS
rose to as great as 2.5 as approached 20. Thus the larger values of
HM/HS
observed in networks perfused with human RBCs could be the result of an
accumulation of RBCs within the network caused by a reduction in the
number of pathways available for flow.
In response to vasodilation, the capillary density at resting (high) PO2 was 150% of that in normal tone networks subjected to bolus infusions of rat RBCs (Table 3) and increased 36% in response to hypoxia. This small increase in capillary density may be caused by the heterogeneous action of papaverine among the various orders of microvessels (2), and thus some vasodilatory reserve remained, although the normal threefold increase in capillary density in response to hypoxia was dramatically attenuated.
In the case of perfusion with the larger human RBCs, capillary density was uniformly diminished by ~35% below that of networks perfused with rat RBCs at all levels of tissue PO2. This attenuation of the number of capillaries perfused was accompanied by an initial 20% rise in TTRBC/TTPl and HM/HS for human versus rat RBCs at high PO2, which was indicative of RBC sequestration in the network. The corresponding fall in these parameters for human RBCs, while TTRBC/TTPl remained elevated above that for rat RBCs, suggests that the stiffer RBCs were redistributed through pathways of lesser resistance, such as thoroughfare channels, as the number of capillaries available for flow increased with hypoxia. Thus, with hypoxia, a smaller percentage of the total vascular volume available for flow can be utilized by the larger, less deformable, human RBCs.
In summary, this study has addressed changes in capillary density in response to tissue hypoxia and the effect of RBC deformability on microvascular function. A significant capillary recruitment was observed in response to hypoxia, which was substantially reduced when the network was vasodilated. Reduced RBC deformability was found to compromise the ability to perfuse these potential additional pathways that ordinarily result from hypoxia. In light of capillary recruitment and a significant increase in flow, the microvascular hematocrit did not show any significant trend, contrary to normal expectations that tissue hematocrit should increase significantly in response to hypoxia. Less deformable RBCs were excluded from the bulk of the capillaries and followed more centralized pathways within the microvasculature. Moreover, these cells also could not take advantage of the attendant capillary recruitment with hypoxia. Thus the ability to recruit additional capillaries by normal compensatory mechanisms with onset of hypoxia may be severely compromised in blood cell disorders manifest by diminished erythrocyte deformability.
| |
ACKNOWLEDGEMENTS |
|---|
The authors thank Karen Trippett for technical assistance with the experiments.
| |
FOOTNOTES |
|---|
This study was funded by National Heart, Lung, and Blood Institute Grants HL-28381 and HL-39286.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests and other correspondence: H. H. Lipowsky, Bioengineering Prog., 233 Hallowell Bldg., Pennsylvania State Univ., University Park, PA 16802 (E-mail: hhlbio{at}engr.psu.edu).
Received 7 October 1998; accepted in final form 25 June 1999.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
1.
Baez, S.
An open cremaster muscle preparation for the study of blood vessels by in vivo microscopy.
Microvasc. Res.
5:
384-394,
1973[Medline].
2.
Ballard, S. T.,
M. A. Hill,
and
G. A. Meininger.
Effect of vasodilation and vasoconstriction on microvascular pressures in skeletal muscle.
Microcirc. Endothelium Lymphatics
7:
109-131,
1991[Medline].
3.
Desjardins, C.,
and
B. R. Duling.
Microvessel hematocrit: measurement and implications for capillary oxygen transport.
Am. J. Physiol.
252 (Heart Circ. Physiol. 21):
H494-H503,
1987
4.
Driessen, G. K.,
C. W. Haest,
H. Heidtmann,
D. Kamp,
and
H. Schmid-Schonbein.
Effect of reduced red cell "deformability" on flow velocity in capillaries of rat mesentery.
Pflügers Arch.
388:
75-78,
1980[Medline].
5.
Gorczynski, R. J.,
and
B. R. Duling.
Role of oxygen in arteriolar functional vasodilation in hamster striated muscle.
Am. J. Physiol.
235 (Heart Circ. Physiol. 4):
H505-H515,
1978
6.
Honig, C. R.,
J. L. Frierson,
and
J. L. Patterson.
Comparison of neural controls of resistance and capillary density in resting muscle.
Am. J. Physiol.
218:
937-942,
1970.
7.
Honig, C. R.,
C. Odoroff,
and
J. Frierson.
Capillary recruitment in exercise: rate, extent, uniformity and relation to blood flow.
Am. J. Physiol.
238 (Heart Circ. Physiol. 7):
H31-H42,
1980
8.
House, S. D.,
and
H. H. Lipowsky.
Microvascular hematocrit and red cell flux in rat cremaster muscle.
Am. J. Physiol.
252 (Heart Circ. Physiol. 21):
H211-H222,
1987
9.
Klitzman, B.,
D. N. Damon,
R. J. Gorczynski,
and
B. R. Duling.
Augmented tissue oxygen supply during striated muscle contraction in the hamster: relative contributions of capillary recruitment, functional dilation, and reduced tissue PO2.
Circ. Res.
51:
711-721,
1982
10.
Klitzman, B.,
and
B. R. Duling.
Microvascular hematocrit and red cell flow in resting and contracting striated muscle.
Am. J. Physiol.
237 (Heart Circ. Physiol. 6):
H481-H490,
1979
11.
Krogh, A.
The number and distributions of capillaries in muscles with calculations of the oxygen pressure head necessary for supplying the tissue.
J. Physiol. (Lond.)
52:
409-415,
1919.
12.
Krogh, A.
The supply of oxygen to the tissues and the regulation of the capillary circulation.
J. Physiol. (Lond.)
52:
457-474,
1919.
13.
Le Noble, J. L.,
G. J. Tangelder,
D. W. Slaaf,
H. van Essen,
R. S. Reneman,
and
H. A. J. Struyker-Boudier.
A functional morphometric study of the cremaster muscle microcirculation in young spontaneously hypertensive rats.
J. Hypertens.
8:
741-748,
1990[Medline].
14.
Lipowsky, H. H.,
L. E. Cram,
W. Justice,
and
M. Eppihimer.
Effect of erythrocyte deformability on in vivo red cell transit time and hematocrit and their correlation with in vitro filterability.
Microvasc. Res.
46:
43-64,
1993[Medline].
15.
Lipowsky, H. H.,
C. B. McKay,
and
J. Seki.
Transit time distributions of blood flow in the microcirculation.
In: Microvascular Mechanics, edited by J. S. Lee,
and T. C. Skalak. New York: Springer, 1989, p. 13-37.
16.
Lipowsky, H. H.,
S. Usami,
and
S. Chien.
Human SS red cell rheological behavior in the micro-circulation of cremaster muscle.
Blood Cells Mol. Dis.
8:
113-126,
1982.
17.
Martin, E. G.,
E. C. Woolley,
and
M. Miller.
Capillary counts in resting and active muscles.
Am. J. Physiol.
100:
407-416,
1932.
18.
Martini, J.,
C. L. Odoroff,
and
C. R. Honig.
Dual effect of oxygen on magnitude and uniformity of coronary intercapillary distance.
Am. J. Physiol.
226:
800-810,
1974.
19.
McKay, C. B.,
and
H. H. Lipowsky.
Arteriovenous distribution of transit times in cremaster muscle of the rat.
Microvasc. Res.
36:
75-91,
1988[Medline].
20.
Meier, P.,
and
K. L. Zierler.
On the theory of the indicator-dilution method for measurement of blood flow and volume.
J. Appl. Physiol.
6:
731-744,
1954
21.
Morff, R. F.
Contribution of capillary recruitment to regulation of tissue oxygenation in rat cremaster muscle.
Microvasc. Res.
36:
150-161,
1988[Medline].
22.
Myers, W. M.,
and
C. R. Honig.
Number and distribution of capillaries as determinants of myocardial oxygen tension.
Am. J. Physiol.
207:
653-660,
1964.
23.
Presson, R. G., Jr.,
J. A. Graham,
C. C. Hanger,
P. S. Godbey,
S. A. Gebb,
R. A. Sidner,
R. W. Glenny,
and
W. W. Wagner, Jr.
Distribution of pulmonary capillary red blood cell transit times.
J. Appl. Physiol.
79:
382-388,
1995
24.
Prewitt, R. L.,
and
P. C. Johnson.
The effect of oxygen on arteriolar red cell velocity and capillary density in the rat cremaster muscle.
Microvasc. Res.
12:
59-70,
1976[Medline].
25.
Reinhart, W. H.,
and
S. Chien.
Roles of cell geometry and cellular viscosity in red cell passage through narrow pores.
Am. J. Physiol.
248 (Cell Physiol. 17):
C473-C479,
1985
26.
Sarelius, I. H.,
D. M. Damon,
and
B. R. Duling.
Microvascular adaptations during maturation of striated muscle.
Am. J. Physiol.
241 (Heart Circ. Physiol. 10):
H317-H324,
1981.
27.
Schmalzer, E. M.,
R. S. Manning,
and
S. Chien.
Filtration of sickle cells: recruitment into a rigid fraction as a function of density and oxygen tension.
J. Lab. Clin. Med.
113:
727-734,
1989[Medline].
28.
Schmid-Schonbein, H.,
and
E. Volger.
Red-cell aggregation and red-cell deformability in diabetes.
Diabetes
25, Suppl. 2:
897-902,
1976.
29.
Schrier, S. L.,
E. Rachmilewitz,
and
N. Mohandas.
Cellular and membrane properties of alpha and beta thalassemic erythrocytes are different: implications for differences in clinical manifestations.
Blood
74:
2194-2202,
1989
30.
Secomb, T. W.,
and
R. Hsu.
Motion of red blood cells in capillaries with variable cross-sections.
J. Biomech. Eng.
118:
538-44,
1996[Medline].
31.
Simchon, S.,
K. M. Jan,
and
S. Chien.
Influence of reduced red cell deformability on regional blood flow.
Am. J. Physiol.
253 (Heart Circ. Physiol. 22):
H898-H903,
1987
32.
Smaje, L.,
B. W. Zweifach,
and
M. Intaglietta.
Micropressures and capillary filtration coefficients in single vessels of the cremaster muscle of the rat.
Microvasc. Res.
2:
96-110,
1970[Medline].
33.
Wayland, N.,
and
P. C. Johnson.
Erythrocyte velocity measurement in microvessels by a two-slit method.
J. Appl. Physiol.
22:
333-337,
1967
34.
Webb, R. C.,
and
J. H. Myers.
Skeletal muscle capillary densities during reactive hyperemia.
Experientia
35:
1476-1477,
1979[Medline].
This article has been cited by other articles:
, Mean
values of both H and TT for rat RBCs; 
, mean values of both H and TT
for human RBCs. Bars represent ±SE for both
PO2 (horizontal), and
HM/HS
and
TTRBC/TTPl
(vertical); no. of observations at each level of
PO2 is shown in parentheses. Dashed
line represents linear regression for rat RBCs, and solid line
represents that for human RBCs.
