Hypoxemia Alters Erythrocyte Perfusion Pattern in the Cerebral Capillary Network

Hypoxemia Alters Erythrocyte Perfusion Pattern in the Cerebral Capillary Network

Microvascular Research 59, 72–79 (2000) doi:10.1006/mvre.1999.2185, available online at http://www.idealibrary.com on Hypoxemia Alters Erythrocyte Pe...

106KB Sizes 0 Downloads 14 Views

Microvascular Research 59, 72–79 (2000) doi:10.1006/mvre.1999.2185, available online at http://www.idealibrary.com on

Hypoxemia Alters Erythrocyte Perfusion Pattern in the Cerebral Capillary Network Ines Krolo and Antal G. Hudetz Department of Anesthesiology, Medical College of Wisconsin, Milwaukee, Wisconsin 53226 Received March 10, 1999

The effect of acute hypoxemia on erythrocyte perfusion rates in individual capillaries of the rat cerebral cortex was studied by intravital video microscopy. The motion of erythrocytes in subsurface capillaries of the frontoparietal cortex was visualized through a closed cranial window using fluorescently labeled red blood cells (FRBC) as markers of flow. FRBC velocity and FRBC supply rate were measured in each capillary at rest, moderate hypoxemia (PaO 2 5 40 mm Hg), and severe hypoxemia (PaO 2 5 26 mm Hg). Lineal density of FRBC in the capillaries was calculated as the ratio of supply rate and velocity. Hypoxemia increased erythrocyte perfusion in virtually all capillaries. Average FRBC supply rate increased by 104% in moderate hypoxemia and by 281% in severe hypoxemia. Average FRBC velocity increased by 66 and 173%, respectively. During severe hypoxemia, FRBC supply rate increased significantly more in capillaries with low resting supply rate compared to those with high resting supply rate. Changes in FRBC velocity exhibited a similar pattern. Lineal density of FRBC increased by 28% in moderate hypoxemia and by 48% in severe hypoxemia. The results suggest that acute hypoxemia promotes perfusion homogeneity and recruitment of erythrocytes in the cerebral capillary network. © 2000 Academic Press

INTRODUCTION Acute systemic hypoxemia is one of the most potent stimuli that elicit a robust increase in cerebral blood

72

flow. The effect of hypoxemia on cerebral capillaries is seen as an increase in flow velocity (Bereczki et al., 1993). Recruitment of previously unperfused capillaries appears to play a minor role in the response (Bereczki et al., 1993; Hudetz et al., 1997). This pattern is consistent with the fact that virtually all microvessels are constantly perfused under normal, resting conditions in the brain (Go¨bel et al., 1989; Villringer et al., 1994; Hudetz et al., 1995). Although cerebral capillaries may not physically open to increase cerebral blood flow, the spatial pattern of cerebral microvascular perfusion may change when blood flow is elevated. This has been observed during hypercapnia (Abounader et al., 1995) and neuronal activation (Vogel and Kuschinsky, 1996). The principal change appears to be an increase in homogeneity of capillary flow that may involve functional recruitment of erythrocytes to dominantly plasmaperfused capillaries (Abounader et al., 1995). Whether a similar redistribution of erythrocyte flow takes place during acute hypoxemia is unclear. The effect of hypoxemia on erythrocyte velocity was previously examined during moderate reductions in arterial PO 2 only (Ivanov et al., 1985; Hudetz et al., 1997). The objective of this study was to determine, using intravital microscopy, the effect of moderate and severe hypoxemia on erythrocyte perfusion in the cerebral capillary network. We measured both linear velocity and supply rate of erythrocytes because these 0026-2862/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

73

Hypoxemia and Cerebral Capillary Flow

parameters may not change by an equal amount when the distribution of erythrocytes in the capillary network is altered. To characterize the heterogeneity of capillary perfusion, we examined the hypoxemic response as a function of resting erythrocyte supply rate and resting erythrocyte velocity. Finally, from the supply rate and velocity of erythrocytes, lineal cell density was calculated. This parameter varies with instantaneous capillary hematocrit and reflects a change in capillary erythrocyte content during hyperemia.

METHODS Experimental Procedures Five adult, male, Sprague–Dawley rats were anesthetized with sodium-pentobarbital (60 mg/kg), tracheotomized, and femoral arterial and venous lines were placed for the measurement of arterial blood pressure, blood gases, and pH and for the infusion of drugs. The head was secured in a stereotaxic apparatus and a closed cranial window was installed over the right parietal cortex using a technique previously described (Hudetz et al., 1995). During the experimental period, the animals were paralyzed with gallamine (25 mg/kg) and artificially ventilated with a mixture of 30% O 2 in N 2. End-tidal CO 2, inspired O 2 concentration, and arterial blood pressure were continuously monitored and maintained at normal values. Anesthesia was maintained with supplemental sodium pentobarbital (7–10 mg/h). Erythrocytes were labeled in vitro with fluorescein isothiocyanate and injected intravenously in tracer quantity (4%). The movement of fluorescent red blood cells (FRBC) in capillary networks approximately 50 mm below the cortical surface was visualized in real time using epifluorescent illumination (100 W mercury) and image intensification. The circulation was studied at 1253 optical magnification (field size: 300 mm 3 400 mm) using a 403 dry objective lens. Following the stabilization of systemic hemodynamic variables, the capillary circulation was videorecorded for 1 min. The fraction of inspired O 2 (FiO 2) was then lowered to 15% for 5 min. The desired O 2

fraction in the tracheal tube was reached in approximately 2 min. An additional 2 min were allowed for the cerebrovascular response and then the microcirculation was videorecorded for 1 min. The FiO 2 was then lowered to 10% for 5 min. Video recording was done during the fifth minute. The resulting arterial O 2 pressures (PaO 2) were 4061 and 26 6 1 mm Hg in moderate and severe hypoxemia, respectively. During hypoxemia, arterial blood pressure was supported by an infusion of the a 1-agonist methoxamine (1.3–3.0 mg/h iv) as used previously (Hudetz et al., 1997). Preliminary studies suggested that 10% FiO 2 was the lowest level at which mean arterial pressure could generally be maintained at its baseline value. The average mean arterial pressures were 119 6 3, 113 6 6, and 108 6 7 mm Hg at rest, moderate hypoxemia, and severe hypoxemia, respectively. Arterial blood samples were taken immediately following each video recording. PaCO 2 was well maintained as reflected by the mean values of 33 6 1, 35 6 2, and 33 6 2 mm Hg in resting condition, moderate hypoxemia, and severe hypoxemia, respectively.

Selection and Classification of Capillaries Capillaries were identified by (a) presence of singlefile FRBC flow, (b) diameter less than 5 mm, (c) resting velocity less than 1.2 mm/s, and (d) organization into an interconnected network of subsurface distribution. In addition, terminal or zero-order capillaries were identified as those segments that carried divergent to convergent flow. Zero-order and first-order capillaries, on both arterial and venous sides, were selected for measurement. Of the 67 capillaries measured, 28 were zero-order capillaries. There was no systematic difference between the responses of first-order and second-order capillary segments. Results from the two capillary types were therefore pooled. The number of capillaries in which cell supply rate was measured varied between 10 and 25 per animal. We attempted to make measurements in all zero- and first-order capillaries visible in the recorded field in each animal. Normal variations in network architecture and image clarity introduced some variance into the number of measurable vessels. No measurable zero- or first-order capillary was excluded.

Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

74

Measurement of FRBC Supply Rate and FRBC Velocity FRBC supply rate and velocity were measured offline from the 1-min long video recordings. FRBC supply rate was measured as the number of labeled cells that passed a reference point in a capillary segment per minute. Supply rate was measured visually, during slow playback of the video tape. A single supply rate per recording period (1 min) was obtained for each capillary. FRBC supply rate was measured in 10 to 25 capillaries in each experiment. FRBC velocity was assessed in a subset of capillaries in each experiment (28 capillaries total). It was more difficult to qualify capillaries for velocity measurement because the latter required a relatively long segment in sharp focus, whereas the measurement of FRBC supply rate could be performed at a single point along each segment. FRBC velocity was measured by tracking the movement of fluorescent cells in several consecutive video fields recorded at a rate of 60 Hz (interlaced video). In each capillary, velocity was measured at approximately 2-s intervals resulting in approximately 30 measurements per capillary. The obtained velocity data were then averaged over the 1-min recording period. Lineal cell density (LCD) was calculated as the ratio of FRBC supply rate and FRBC velocity and was expressed in units of fluorescent cells per millimeter length of capillary.

Statistical Analysis All measured data were tested for normality using the Liliefors (one-way Kolmogorov–Smirnov) test. When this criterion was met, parametric statistical tests were used. Otherwise, nonparametric tests were used. FBRC supply rates during rest, moderate hypoxemia, and severe hypoxemia were compared using the Friedman statistic and, after logarithmic transformation of the data, by repeated measures ANOVA and Scheffe’s test. The effects of moderate and severe hypoxemia on FRBC supply rate were compared using the Wilcoxon signed-rank test. Differences in FRBC velocity obtained during rest and two levels of hypoxemia were tested for significance using repeated measures ANOVA followed by the Student–Newman–

Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

Krolo and Hudetz

Keuls test. For severe hypoxemic data only the unpaired t test was used. Multiple comparison of LCD data was performed using the Student–Newman– Keuls test following ANOVA and after subtracting the mean in each capillary to reduce between-subject variance.

RESULTS Red Cell Supply Rate at Rest and Hypoxemia Under resting conditions, the 1-min average supply rate of FRBC varied from capillary to capillary between 0 and 497 cells per minute. Ninety percent of the capillaries had a rate of at least 5 labeled cells/min. The distribution of supply rate data was significantly different from normal. To describe the central tendency of data we applied a logarithmic transformation. Figure 1 illustrates the distribution of measured FRBC supply rates before and after transformation. After transformation, a small population of capillaries with very low, potentially intermittent, flow could be isolated (left tail of the transformed distribution). These capillaries represented approximately 10% of the total population. Intermittence of flow could not be verified due to the low fraction of labeled cells (4%). Statistical distribution of supply rate data from the well-perfused capillaries was not different from the normal distribution. The antilog of the mean of this distribution, equivalent to the most frequent supply rate (mode) of the original distribution, was 70 labeled cells per minute. Hypoxemia produced an increase in FRBC supply rate in all except two capillaries. In further two capillaries, FRBC flow could be detected during hypoxemia only. In moderate hypoxemia, the mode of supply rate distribution increased to 106 cells/min, and in severe hypoxemia, to 198 cells/min. The amount of increase in FRBC supply rate was significantly greater in severe than in moderate hypoxemia (281 vs 104%, respectively, P , 0.01). Next we examined whether the hypoxemic response of FRBC supply rate showed any difference as a function of the resting supply rate. To this end,

75

Hypoxemia and Cerebral Capillary Flow

Red Cell Velocity at Rest and Hypoxemia Average FRBC velocity at rest was 0.41 6 0.21 (SD) mm/s. The velocity data were normally distributed. As expected, FRBC velocity tended to be greater in first-order capillary segments than in zero-order segments (0.52 vs 0.35 mm/s, respectively, P , 0.05). However, there was no difference in the velocity response to hypoxemia between the two orders of capillaries. Average FRBC velocity increased to 0.68 6 0.31 mm/s in moderate hypoxemia and to 1.12 6 0.52 in severe hypoxemia. The corresponding relative increases were 66 and 173%, respectively. Both changes were significant (P , 0.01). The velocity response to hypoxemia as a function of resting velocity was also examined. As with the FRBC supply rate, the percentage increase in velocity correlated negatively with the resting velocity in severe but not in moderate hypoxemia, suggesting that the slower capillaries exhibited a significantly greater velocity response. Because the correlation coefficient was not too high (r 5 0.59), we tested this prediction by another method. The velocity data were separated into FIG. 1. Frequency distribution of fluorescent red blood cell (FRBC) supply rates measured in 67 capillaries of 5 rats. (A) The distribution can be described as the Weibull distribution (r 5 0.95, P , 0.05). (B) After logarithmic transformation, the central tendency of the data is more evident. Approximately 10% of the capillaries exhibit distinctly low supply rates that could be due to intermittent perfusion of these vessels. The greater part of the transformed distribution is not different from normal (P 5 0.28).

the relative increase in FRBC supply rate in each capillary was plotted against the resting supply rate in the same capillary. The supply rate increase was calculated as (SR 2 SR R )/SR SH p100 where SR stands for the actual supply rate and SR R and SR SH stand for supply rates at rest and severe hypoxemia, respectively, in the same capillary. As seen in Fig. 2, a negative correlation (r 5 0.68) between the change in supply rate and the resting supply rate was present in severe but not in moderate hypoxemia. These results indicate that in severe hypoxemia, FRBC supply rate increases were greater in capillaries with low resting supply rate than in those with high resting supply rate.

FIG. 2. Hypoxemia-induced increase in FRBC supply rate as a function of resting supply rate in the same capillary. Supply rate increases are plotted as a percentage of the severe hypoxemic value. Full circles, moderate hypoxemia; open circles, severe hypoxemia. A significant negative correlation is observed in severe but not moderate hypoxemia.

Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

76

Krolo and Hudetz

In moderate hypoxemia, LCD increased by 28% and in severe hypoxemia by 48%. The differences in LCD among rest, moderate, and severe hypoxemia were all significant (P , 0.01, Fig. 4). Also, the increase in LCD from rest to severe hypoxemia was significantly greater than the increase from rest to moderate hypoxemia (P , 0.01). The change in LCD did not correlate with resting LCD, resting supply rate, or resting velocity.

DISCUSSION

FIG. 3. Severe hypoxemia-induced increases in FRBC velocity as a function of resting velocity. The two symbols indicate two groups of capillaries separated near the median of the resting velocity. The velocity increase is plotted as a percentage of the hypoxemic velocity. Note that capillaries with low resting velocity (below median, triangles) exhibit a significantly greater (P , 0.01) response to severe hypoxemia than those with high resting velocity (above median, rectangles).

two groups near the median of the resting velocity and the group means of percentage velocity change were compared by the unpaired t-test. As illustrated in Fig. 3, the velocity response was significantly greater in capillaries with low resting velocity than in those with high resting velocity.

This work was undertaken to determine the effect of acute hypoxemia on erythrocyte perfusion pattern in the cerebrocortical capillary network. The study advances our previous work (Hudetz et al., 1997) on the same subject in several aspects. First, we investigated the effect of both moderate and severe hypoxemia. Only moderate hypoxemia was studied before. We suspected that severe hypoxemia may elicit a qualitatively different response from that produced by moderate hypoxemia because such a difference had been observed in experiments with moderate and severe hypercapnia (Hudetz et al., 1997). Second, we mea-

Effect of Hypoxemia on Lineal Cell Density Lineal cell density of labeled erythrocytes was calculated from corresponding measured values of FRBC supply rate and FRBC velocity in each capillary. Average LCD at rest was 1.45 cells/mm which, at a labeled fraction of 4%, yielded a true red cell density estimate of 36 cells/mm. This value predicts that the erythrocytes travel in cerebral capillaries approximately two cell lengths apart. This result reflects an average trend, as LCD may vary in time as well as from capillary to capillary, indicated by a coefficient of variation of 0.73.

Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

FIG. 4. Frequency distributions of lineal cell density (LCD) of FRBC in the capillary during rest, moderate hypoxemia, and severe hypoxemia. A progressive shift of the distribution to the right is evident with increasing severity of hypoxemia and is reflected in significant increases of the mean. Data are from 28 capillaries of 5 rats.

Hypoxemia and Cerebral Capillary Flow

sured both supply rate and velocity of erythrocytes. We did this to account for the possibility that supply rate may not change by the same amount as velocity if erythrocyte distribution or hematocrit were altered. Finally, we classified the capillary segments according to their resting flow rate and examined the hypoxemic response as a function of this variable. The major findings of this study are that in hypoxemia, red cell supply rate and velocity increased in practically all capillaries and that in severe hypoxemia, supply rate and velocity increased more in capillaries with low resting values of these parameters. In addition, calculated lineal density of erythrocytes increased in both moderate and severe hypoxemia. To date, few studies have directly measured capillary flow in the brain during rest and hypoxemia. Ivanov et al. (1985) measured the velocity of flow in capillaries of the rat cerebral cortex using darkfield microscopy and found a 66% increase in velocity at a PaO 2 of 38 mm Hg. Our present result is identical with this datum at a similar PaO 2. This velocity change is, however, notably greater than the 35% value reported previously (Hudetz et al., 1997). This difference may be due to the experimental design, particularly the selection of capillaries. Zero and first order capillaries tend to have the lowest resting velocity and may exhibit a greater relative increase in velocity than a random population of capillaries studied previously. Finally, the maximum increase in FRBC supply rate at 280% was substantially greater than that reported for a corresponding increase in cerebrocortical blood flow (Bereczki et al., 1993) at a comparable PaO 2. This interesting difference may be due to the higher selectivity of our measurement for true capillaries and/or to a difference between erythrocyte flow and whole blood flow in the cerebral microcirculation.

Increased Homogeneity of Erythrocyte Perfusion in Hypoxemia One of the principal questions to be answered by this study was whether red cell velocity and supply rate were altered nonuniformly across the capillary network in acute hypoxemia. Although FRBC velocity and supply rate increased in virtually all capillaries, the magnitude of these increases were variable and, in

77

severe hypoxemia, appeared to depend on the resting velocity or supply rate. Since there was a greater increase in capillaries with low resting flow than in those with high resting flow (either FRBC velocity or supply rate), the overall distribution of erythrocyte flow had to become more homogeneous in severe hypoxemia. This conclusion is consistent with previous reports of increased homogeneity of perfusion of the cerebral capillary bed in hypercapnia (Abounader et al., 1995; Villringer et al., 1994; Hudetz et al., 1997) and functional hyperemia (Vogel and Kuschinsky, 1996). A generalization of these findings thus leads to the proposition that the perfusion of cerebral capillaries by erythrocytes becomes more homogeneous when a sufficient need to enhance transcapillary exchange is present. The possibility that pentobarbital anesthesia disturbed the distribution of capillary flow has to be addressed. One might speculate that if pentobarbital produced a nonuniform decrease in capillary flow, then this could have led to a greater response to hypoxemia in the slower capillaries which could be fewer in the conscious animal. Bereczki et al. (1993) compared the velocities of red cell and plasma flows in awake versus pentobarbital-anesthetized rats and found that while both velocities were lowered, the percentage of erythrocyte-perfused capillaries was not decreased by pentobarbital. They found that red cell space and microvessel hematocrit increased, possibly due to recruitment of red-cell-perfused capillaries or to the Fahraeus effect. Nevertheless, an increase in homogeneity of capillary perfusion in hypercapnia was found in both pentobarbital-anesthetized (Hudetz et al., 1997) and conscious rats (Abounader et al., 1995). Therefore, it is unlikely that the nonuniform response of capillary flow to hypoxemia was an artifact of pentobarbital. With our technique, RBC flow in capillaries of the first cortical layer is assessed. Capillary density and preferential orientation vary in different layers of the cerebral cortex (Dunning and Wolff, 1937; Duvernoy et al., 1981). In addition, layer I is sparse in cell bodies but rich in dendrites and synapses which entail significant metabolic activity. The magnitude and time course of blood flow changes may be different in various cortical layers (Moskalenko et al., 1997).

Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

78

Whether hypoxemia produces a different pattern of capillary flow response in different cortical layers is unknown. Extrapolation of our results to deeper regions of cerebral gray matter is nevertheless encouraged by the fact that homogenization of capillary perfusion in hypercapnia has been seen in multiple regions of the brain (Abounader et al., 1995).

Erythrocyte Recruitment in Hypoxemia In this study, hypoxemia was accompanied by a greater increase in FRBC supply rate than in FRBC velocity. Consistently, the ratio of these parameters, equivalent to LCD, the lineal density of FRBC in the capillary, was increased. The latter finding suggests that in hypoxemia the number of erythrocytes traversing the capillary at any moment was increased, a phenomenon that may be called erythrocyte recruitment. The term “erythrocyte recruitment” should be contrasted with the term “capillary recruitment.” Capillary recruitment is defined as the opening of previously unperfused capillaries with an increase in overall blood flow (Kuschinsky and Paulson, 1992). Capillary recruitment appears to make no or minor contribution to cerebral hyperemia (Go¨bel et al., 1989; Bereczki et al., 1993; Villringer et al., 1994; Hudetz et al., 1997, etc.). Consistently, we found only 2 of the 67 measured capillaries that showed transition from zero to nonzero FRBC supply rate during hypoxemia. Some investigators have suggested that cerebral capillaries are continuously perfused with plasma but experience intermittence in erythrocyte flow (Kislyakov et al., 1987; Mchedlishvili et al., 1997). In our study, 10% of the capillaries had low, potentially intermittent flow at rest. Three of these capillaries had a supply rate of 1 FRBC per minute that, at a labeled cell fraction of 4%, predicts a supply rate of 1 erythrocyte every 2 s. As 2 s is about twice the capillary transit time, these capillaries may have been devoid of erythrocytes half of the time. On the other hand, 90% of the capillaries were perfused at a rate between 5 and 500 FRBC per minute implying essentially continuous perfusion by erythrocytes. It appears therefore that in most of the capillaries, hypoxemia-induced increases in red cell flow are graded and that this graded change

Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

Krolo and Hudetz

in flow determines the pattern of erythrocyte perfusion of the cerebral microvascular network. A limitation of the labeled cell technique that may have affected our results is that erythrocyte velocity could only be measured when a labeled cell was present in the capillary. This restriction may have resulted in overestimation of the average velocity and LCD if erythrocyte perfusion was intermittent. (The measurement of FRBC supply rate is not affected.) This means that if the intermittence was reduced in hypoxemia, LCD could have increased less than predicted by the labeled cell technique. Nevertheless, since intermittent flow occurs in a small fraction of capillaries, the hypoxemic changes in average velocity and LCD should be close to those determined here.

Physiological Mechanisms and Implications The physiological mechanism that underlies the observed change in erythrocyte perfusion pattern is currently unclear. Capillary red cell content as indicated by LCD increased significantly in both moderate and severe hypoxemia whereas a change in perfusion homogeneity could be detected in severe hypoxemia only. This difference suggests that the underlying mechanism is complex. We speculate that erythrocyte recruitment could be due to an increase in flow cross section of the capillary. Increases in cerebral capillary diameter have been reported in hypercapnia (Atkinson et al., 1990; Duelli and Kuschinsky, 1993; Villringer et al., 1994) and may also be present in hypoxemia (Bereczki et al., 1993). As more erythrocytes enter the capillaries, tube hematocrit and flow resistance may rise in the high flow vessels and divert erythrocytes to poorly perfused capillaries. Shear-dependent variations in the capillary endothelial macromolecular layer may contribute to these effects (Pries et al., 1997). The functional significance of increased perfusion homogeneity is currently unclear but may be related to potentially more efficient transcapillary exchange in the capillary network. In summary, our intravital microscopic study of the rat cortical microcirculation reveals significant, graded increases in erythrocyte supply rate, velocity, and cell density in essentially all capillaries during acute hypoxemia. The distribution of erythrocyte perfusion in

Hypoxemia and Cerebral Capillary Flow

the capillary network becomes more homogeneous during severe hypoxemia and is paralleled by an increase in capillary red cell content, i.e., by erythrocyte recruitment.

ACKNOWLEDGMENTS This work was supported in part by grants from the National Science Foundation, BES-9411631, the American Heart Association, GIA-95009340, and the National Institute of Health, GM-56398. The technical assistance from James D. Wood and the secretarial assistance and from Anita Tredeau and Angela Barnes are greatly appreciated.

REFERENCES Abounader, R., Vogel, J., and Kuschinsky, W. (1995). Patterns of capillary plasma perfusion in brains of conscious rats during normocapnia and hypercapnia. Circ. Res. 76, 120 –126. Atkinson, J. L. D., Anderson, R. E., and Sundt, T. M. (1990). The effect of carbon dioxide on the diameter of brain capillaries. Brain Res. 517, 333–340. Bereczki, D., Wei, L., Otsuka, T., Acuff, V., Pettigrew, K., Patlak, C., and Fenstermacher, J. (1993). Hypoxia increases velocity of blood flow through parenchymal microvascular systems in rat brain. J. Cereb. Blood Flow Metab. 13, 475– 486. Duelli, R., and Kuschinsky, W. (1993). Changes in brain capillary diameter during hypocapnia and hypercapnia. J. Cereb. Blood Flow Metab. 13, 1025–1028. Dunning, H. S., and Wolff, H. G. (1937). The relative vascularity of various parts of the central and peripheral nervous system of the cat and its relation to function. J. Comp. Neurol. 67, 433– 450.

79 Duvernoy, H. M., Delon, S., and Vannson, J. L. (1981). Cortical blood vessels of the human brain. Brain Res. Bull. 7, 519 –579. Go¨bel, U., Klein, B., Schro¨ck, and Kuschinsky, W. (1989). Lack of capillary recruitment in the brains of awake rats during hypercapnia. J. Cereb. Blood Flow Metab. 9, 491– 499. Hudetz, A. G., Biswal, B. B., Fehe´r, G., and Kampine, J. P. (1997). Effects of hypoxia and hypercapnia on capillary flow velocity in the rat cerebral cortex. Microvasc. Res. 54, 35– 42. Hudetz, A. G., Fehe´r, G., Weigle, C. G. M., Knuese, D. E., and Kampine, J. P. (1995). Video microscopy of cerebrocortical capillary flow: Response to hypotension and intracranial hypertension. Am. J. Physiol. 268, H2202–H2210. Ivanov, K. P., Kalinina, M. K., and Levkovich, Yu. I. (1985). Microcirculation velocity changes under hypoxia in brain, muscles, liver, and their physiological significance. Microvasc. Res. 30, 10 – 18. Kislyakov, Yu. Ya., Levkovitch, Yu. I., Shuymilova, T. E., and Vershinina, E. A. (1987). Blood flow fluctuations in cerebral cortex microvessels. Int. J. Microcirc. Clin. Exp. 6, 3–13. Kuschinsky, W., and Paulson, O. B. (1992). Capillary circulation in the brain. Cerebrovasc. Brain Metab. Rev. 4, 261–286. Mchedlishvili, G., Varazashvili, M., Mamaladze, A., and Momtselidze, M. (1997). Blood flow structuring and its alterations in capillaries of the cerebral cortex. Microvasc. Res. 53, 201–210. Moskalenko, Y. E., Wooley, T. A., Rovainen, C. M., Weinstein, G. B., Erinjeri, J., Liu, D., Semernia, V. N., and Mitrofanov, V. F. (1997). LCBF in different cortical layers of rat barrel cortex during whisker stimulation. J. Cereb. Blood Flow Metab. 17(Suppl. 1), S601. Pries, A. R., Secomb, T. W., Jacobs, H., Sperandio, M., Osterloh, K., and Gaehtgens, P. (1997). Microvascular blood flow resistance: Role of endothelial surface layer. Am. J. Physiol. 273, H2272– H2279. Villringer, A., Them, A., Lindauer, U., Einha!upl, K., and Dirnagl, U. (1994). Capillary perfusion of the rat brain cortex: An in vivo confocal microscopy study. Circ. Res. 75, 55– 62. Vogel, J., and Kuschinsky W. (1996). Decreased heterogeneity of capillary plasma flow in the rat whisker-barrel cortex during functional hyperemia. J. Cereb. Blood Flow Metab. 16, 1300 –1306.

Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.