JULIAN H. LOMBARD AND BRIAN R. DULING

Transcription

1 546 Relative Importance of Tissue Oxygenation and Vascular Smooth Muscle Hypoxia in Determining Arteriolar Responses to Occlusion in the Hamster Cheek Pouch JULIAN H. LOMBARD AND BRIAN R. DULING SUMMARY Cheek pouches of anesthetized hamsters were suffused with solutions with low, medium, and high oxygen tension, and arteriolar diameter changes were measured during and after occlusions of either single arterioles or the whole vascular bed. PO 2 was measured with microelectrodes on arteriolar walls and at surrounding tissue sites to indicate oxygen availability to the vascular smooth muscle and parenchymal cells, respectively. In each suffusion solution, whole pouch occlusion (WPO) and microvessel occlusion (MVO) induced quantitatively similar changes in tissue and periarteriolar PO 2 ; however, WPO resulted in larger diameter increases and longer recovery times than MVO. High suffusion solution PO 2 attenuated the reduction in tissue and periarteriolar PO 2 during occlusion (WPO and MVO), reduced the dilation during occlusion (WPO and MVO), and was associated with faster recovery of arteriolar diameters following release of WPO. Postocclusion recovery of periarteriolar PO 2 and tissue PO 2 in low oxygen suffusion was significantly faster than that of arteriolar diameters (WPO and MVO). Following WPO, periarteriolar PO 2 often had fully recovered while diameter was at its peak value. During intermediate and high PO 2 suffusion, arteriolar dilation occurred despite relatively high periarteriolar oxygen tensions. The evidence suggests that arteriolar responses to occlusion are determined primarily by indirect mechanisms, i.e., those mediated through parenchymal cell metabolites, rather than by the direct effects of oxygen deficiency on the vascular smooth muscle cells of the arteriolar media. Reactive hyperemia is the period of increased blood flow following occlusion of the blood supply to a vascular bed. Two of the major mechanisms proposed to explain this response involve oxygen either by its effect on vascular smooth muscle or by an indirect mechanism mediated through metabolism of the parenchymal cells. The indirect mechanism involves dilation of the small resistance vessels induced by the accumulation of vasoactive tissue metabolites during the period of occlusion, 1 whereas the direct effect refers to relaxation of the vascular smooth muscle induced by low PO 2 in the muscle cells. 2 Most studies of reactive hyperemia have been performed on whole vascular beds with only indirect indications of the precise behavior of the resistance vessels. Microcirculatory studies have the distinct advantage of permitting direct access to the small blood vessels and observation of their responses to various stimuli. However, relatively few such investigations of reactive hyperemia have been conducted. 3 " 5 Since oxygen availability to tissues and to the microcir- From the Department of Physiology, University of Virginia School of Medicine, Charlottesville, Virginia. Supported by National Institutes of Health Grant HL This work was conducted during Dr. Duling's tenure as an Established Investigator of the American Heart Association. Dr. Lombard was supported by National Institutes of Health Postdoctoral Fellowship HL Address for reprints: Dr. Julian H. Lombard, Department of Physiology, University of Virginia School of Medicine, Charlottesville, Virginia Preliminary reports of this work were presented at the First World Conference of Microcirculation, Toronto, Canada, June 1975; and at the Microcirculatory Society Meeting, Anaheim, California, April Received November ; accepted for publication March 1, culation can be altered by varying the PO 2 of a suffusion solution covering the tissue, suffused microvessel preparations permit an evaluation of the importance of oxygen in the vascular response during both occlusion and the ensuing period of reactive hyperemia. Oxygen tensions, measured with microelectrodes on the walls of arterioles and at various tissue sites, are good indices of oxygen availability to the vascular smooth muscle and parenchymal cells, respectively. 6 The object of the present study was to use these methods to determine whether the arteriolar dilation which occurs in response to vascular occlusion correlates more closely with oxygen deficiency at the vascular smooth muscle or with oxygen deficiency of the parenchymal tissue. Methods Male golden hamsters ( g) were anesthetized with an intraperitoneal injection of sodium pentobarbital, 60 mg/kg. A tracheostomy was performed to ensure a patent airway, and a femoral vein was cannulated. Supplemental anesthesia was administered, iv, as required. Physiologic saline solution was infused at a rate of 0.42 ml/hr to compensate for evaporative water loss and to maintain circulating blood volume. Preparation of a single layer suffused cheek pouch was performed as previously described. 7 The blood supply to the whole vascular bed was occluded by a lifting ligature around the common carotid artery or by compression of the base of the pouch with a hydraulic occluder 8 or a soft silicone rubber snare. Single arterioles were occluded by compression with glass micropipettes guided by a Leitz or De Fonbrune micromanipulator.

2 ROLE OF OXYGEN DURING OCCLUSION/Lombard and Duling 547 The hamster was placed on the stage of a Leitz Labolux II microscope equipped with a 50 x UMK objective and the preparation was transilluminated with a high intensity xenon lamp filtered at 540 nm. The cheek pouch was continuously suffused with a bicarbonate-buffered physiologic saline solution, ph 7.35, with the following millimolar composition: NaCl, 131.9; KC1, 4.7; CaCl 2, 2.0; MgSO 4,1.17; and NaHCO 3,20.0. The temperature of the pouch was maintained at C and the preparation was allowed to equilibrate for each experiment until diameters and oxygen tensions were stable. The presence of tone in arterioles under observation was verified by observing the dilation produced by the topical application of a 10~ 4 M solution of adenosine or acetylcholine. The microcirculation was also observed on a Cohu closed circuit television system having a useful magnification of about 500x. Inside diameters of arterioles were measured continuously before, during, and after occlusion with a pair of movable lines generated by a Colorado Video Analyzer (model 321). The DC voltage output from the Video Analyzer was recorded by a Brush model 260 chart recorder. The entire system was calibrated against a Vickers image-shearing eyepiece, and had an overall accuracy of ±1 ^m. Oxygen tensions on the walls of arterioles and in the tissue were measured amperometrically with oxygen microcathodes 9 as described previously. 6 The tissue sites selected were free of capillaries and thus reflect minimal PO 2 in the tissue. The microelectrodes were also used to measure the PO 2 of the suffusion solution as it flowed over the pouch. Electrodes were calibrated in saline equilibrated with 100% N 2 and room air before and after each experiment, and readings were discarded if any significant change occurred in the calibration currents. Currents, measured on a Keithley model 602 picoammeter, were in the range of 0.1 x 10"" A in 0% O 2, and the ratio of 0% O 2 /21 % O 2 currents was 0.05 ± (SEM). The tip size of the microelectrodes used in this study ranged from 1 to 4 /am. To determine the effect of oxygen supply on the responses of arterioles to occlusion, the oxygen content of the suffusion solution was altered by equilibrating the solution in the supply reservoir with gas mixtures containing either 0% O 2, 5% O 2, or 10% O 2, with 5% CO 2 and N 2 as a filler gas. The PO 2 of the various suffusion solutions, measured over the pouch with microelectrodes, was 7.2 ± 1.3 mm Hg for the 0% O 2 solution, 35.0 ±1.1 mm Hg for the 5% O 2 solution, and mm Hg for the 10% O 2 solution. Several parameters of the microvascular response were evaluated in an effort to develop expressions relevant to the pressure-flow relations in whole organs. Measurement of diameter and changes in diameter at the end of an occlusion period were used as corollaries to changes in peak flow. Similarly, the time required for recovery of the arteriolar diameter following occlusion should be an index of the duration of reactive hyperemia. We have therefore expressed the data as the half-time for recovery. Results ARTERIOLAR DIAMETER CHANGES DURING OCCLUSION Table 1 compares the final diameter increases (/*m) following 60 seconds of whole pouch occlusion or single vessel occlusion during suffusion with solutions with low, intermediate, and high oxygen content. In the case of single vessel occlusions, all measurements were made downstream from the pipette to ensure that intravascular pressure decreased during the occlusion. Diameter increases during single arteriole occlusions have also been described in detail as part of a previous investigation. 10 Arteriolar diameters increased during occlusion with all three suffusion conditions and, in each solution, diameter increases during whole bed occlusion were significantly greater than those during single vessel occlusion. During both single vessel and whole pouch occlusion, less dilation occurred as the PO 2 of the suffusion solution was increased. RECOVERY TIMES FOLLOWING WHOLE POUCH AND SINGLE VESSEL OCCLUSIONS Table 2 compares the recovery times of arteriolar diameter following release of whole bed and single vessel occlusion during 0% O 2, 5% O 2, and 10% O 2 suffusion. The time necessary for 50% recovery from the maximum dilation attained during the occlusion was measured as an index of the recovery rate. During all suffusions, the half- TABLE 1 Peak Diameter Increase during 1-Minute Occlusions of Single Arterioles or the Whole Vascular Bed Suffusion 0%O 2 5% Oj 10% O, Control Diameter increase (jim) 20.7 ± ± 0.6 (24) [17] 17.7 ± ± 0.6 (12) [9] 15.1 ± ± 0.5t (15) [11] Whole pouch occlusion Diameter increase Control (/im) 21.7 ± ± 1.5* (24) [15] 18.6 ± ± 1.0* (8) [5] 18.8 ± ± 1.2't (13) [7] All values are expressed as mean ± SEM. Parentheses indicate numbers of arterioles observed; brackets indicate the number of animals. * P < 0.05 vs. single vessel occlusion at same suffusion solution PO 2. t P < 0.05 vs. same type of occlusion during 0% O 2 suffusion.

3 548 CIRCULATION RESEARCH VOL. 41, No.'4, OCTOBER 1977 TABLE 2 Half-Times of Arteriolar Diameter Recovery following 1-Minute Occlusions of Single Arterioles or the Whole Vascular Bed Suffusion 0%O 2 5% O 2 10% O ± 1.4 (27) [23] 6.3 ± 1.4 (6) [6] 10.7 ± 2.6 (12) [7] Time (sec) Whole bed occlusion 33.0 ± 3.8* (36) [24] 18.6 ± 2.5*t (16) [8] 17.6 ± 2.6*t (23) [14] All values are expressed as the time (mean ± SEM) from release of occlusion until 50% recovery from the maximum diameter attained during 1-minute occlusions of either single arterioles or the whole vascular bed. Parentheses indicate numbers of arterioles observed; brackets indicate the number of animals. * P < 0.05 vs. single vessel occlusion at same suffusion solution PO 2. t P < 0.05 vs. whole bed occlusion in 0% O 2 suffusion. time of recovery following whole pouch occlusion was significantly longer than that following single vessel occlusion. Recovery times following 1-minute whole pouch occlusions were reduced as the PO 2 of the suffusion solution was increased. Recovery times following 1-minute occlusions of single arterioles were not significantly influenced by the oxygen content of the suffusion solution. Following 3-minute single vessel occlusions, recovery of arteriolar diameters were slower in low oxygen suffusion. The halftime of recovery following 3-minute single vessel occlusions in 0% O 2 suffusion averaged 22.2 ± 4.1 seconds (n = 19), compared with half-times of 11.5 ± 1.4 seconds for 1-minute occlusions in 0% O 2 suffusion (n = 27) (P < 0.05) and 11.8 ± 3.3 seconds for 3-minute occlusions in 10% O 2 suffusion (n = 10) (P < 0.05). PERIARTERIOLAR PO 2 DURING SINGLE VESSEL AND WHOLE BED OCCLUSION Table 3 compares changes in periarteriolar PO 2 during whole bed occlusion and during single vessel occlusion under the various suffusion conditions. With the low oxygen suffusion, periarteriolar PO 2 reached similar minimum values of 9.0 ± 2.1 mm Hg during single vessel occlusion and 7.5 ± 1.9 mm Hg during whole bed occlusion. PO 2 on the arteriolar wall reached 0 mm Hg during two of 13 single vessel occlusions and during two of 10 whole bed occlusions. Periarteriolar PO 2 also fell below 2 mm Hg during two other single vessel occlusions and one other whole bed occlusion. When the occlusion was released, oxygen tension on the arteriolar wall rapidly recovered to a value almost identical to the preocclusion PO 2 (29.5 ± 1.8 mm Hg following single vessel occlusion and 26.1 ± 2.4 mm Hg following whole pouch occlusion). During single arteriole occlusion in 5% O 2 suffusion, periarteriolar PO 2 again declined markedly, but reached a plateau at a minimum value of 18.4 ± 1.9 mm Hg. This is significantly higher than the minimum value observed during 0% O 2 suffusion (P < 0.01). PO 2 on the arteriolar wall failed to fall below 8 mm Hg during any occlusion. Periarteriolar PO 2 changes in response to occlusion of the whole vascular bed were similar to those observed during single arteriole occlusion. During 10% O 2 suffusion, periarteriolar PO 2 tended to decrease slightly during both single arteriole and whole bed occlusion. When evaluated by Student's /-test for paired samples, this decrease was significant (P < 0.02) only during whole bed occlusion. The average minimum periarteriolar PO 2 of 28.4 ±4.3 mm Hg (single arteriole occlusion) and of 27.7 ±2.2 mm Hg (whole bed occlusion) were not significantly different from each other, but were significantly higher than those occurring during single vessel and whole bed occlusions with 0% O 2 suffusion (P < 0.001). TISSUE PO 2 DURING SINGLE VESSEL AND WHOLE BED OCCLUSION Our measurements of tissue PO 2 under conditions of unrestricted blood flow are in good agreement with those previously described by others. 6 ' "~ 15 The changes in tissue PO 2 during occlusion on 0%, 5%, and 10% O 2 suffusion are compared in Table 4. Tissue PO 2 changes were similar in response to single vessel and whole bed occlusion. With 0% O 2 suffusion, tissue PO 2 reached 0 mm Hg in TABLE 3 Periarteriolar P0 2 Changes in Response to 1-Minute Occlusions of Single Arterioles or the Whole Vascular Bed Periarteriolar PO, changes (mm Hg) Whole pouch occlusion Suffusion Conlrol Minimum Control Minimum 0% O ± ± 2.1 5% O 2 (13)[8] 33.1 ± ± 1.9* 10% O 2 (11) [ ± ± 4.2't (7) [5] 26.3 ± ± 1.9 (10) [7] 32.8 ± ± 1.4* (8) [4] 32.0 ± ± 2.3* (8) [6] All values are expressed as mean ± SEM. Control values represent PO 2 immediately prior to occlusion, and minimum values refer to the lowest PO, reached during the occlusion. Parentheses indicate number of arterioles observed; brackets indicate the number of animals. * P < 0.05 vs. same type of occlusion in 0% O 2 suffusion. t P < 0.05 vs. same type of occlusion in 5% O 2 suffusion.

4 ROLE OF OXYGEN DURING OCCLUSION/Lombard and Duling 549 p<j>01 A (12) p<01 B «) Whole Pouch Occlusion A-Periarteriolar POj B-Tissue PO2 C- Diameter (19) Single Arteriole Occlusion FIGURE 1 Comparison of recovery rates of periarteriolar PO 2 (A), tissue PO 2 (B), and arteriolar diameter (C) following 1- minute occlusions of single arterioles or the whole vascular bed during 0% O 2 suffusion. Data are plotted as the elapsed time in seconds (mean ± SEM) from release of occlusion until 50% recovery. Significance levels refer to comparison of recovery times of periarteriolar PO 2 (A) or tissue PO 2 (B) with the recovery time for ateriolar diameter (C). Numbers refer to the number of arterioles observed. With 10% O 2 suffusion, tissue PO 2 exhibited a small but significant decrease (Student's /-test for paired samples) in response to both single vessel and whole pouch occlusion. The minimum tissue PO 2 reached during whole bed and that during single vessel occlusion were not significantly different from each other, but were significantly higher than those during 0% and 5% O 2 suffusion. Preocclusion values of tissue PO 2 during 10% O 2 suffusion were also significantly greater than those during 0% O 2 suffusion (P < 0.001) and 5% O 2 suffusion (P < 0.05). TIME COURSE OF DIAMETER AND PO 2 RECOVERY FOLLOWING OCCLUSION (0% O 2 SUFFUSION) Figure 1 compares the half-times of recovery of arteriolar diameter, periarteriolar PO 2, and tissue PO 2 following single arteriole occlusion and whole pouch occlusion in low oxygen suffusion. Periarteriolar PO 2 recovered significantly faster than arteriolar diameter, and, following whole bed occlusion, vascular diameters were often at their peak value for some time after periarteriolar PO 2 had fully recovered. After release of both single arteriole occlusion and whole pouch occlusion, tissue PO 2 recovered more slowly than periarteriolar PO 2, but significantly faster than arteriolar diameter. two of 10 single vessel occlusions and in five of nine whole bed occlusions. Tissue PO 2 also fell below 2 mm Hg during seven of the single vessel occlusions and during three of the whole bed occlusions. During whole bed occlusion, tissue PO 2 tended to reach lower minimum values than during single vessel occlusion, but this difference was not significant. Upon release of occlusion, tissue PO 2 recovered rapidly with a half-time of 3.6 ± 0.4 seconds following single arteriole occlusion (n = 11) and 10.3 ± 2.2 seconds following whole bed occlusion (n = 8). The half-time of recovery following single vessel occlusion was significantly faster than that following whole bed occlusion (P < 0.005). In 5% O 2 suffusion, tissue PO 2 also declined sharply during occlusion, reaching minimum values which were intermediate to those reached during occlusion in the 0% O 2 and 10% O 2 suffusions. Upon release of occlusion, tissue PO 2 again returned rapidly to control values. Discussion The present experiments indicate that the magnitude and duration of arteriolar dilation in response to occlusion are not under direct and continuous control by oxygen deficiency at the level of the vascular smooth muscle cells. After release of both whole bed and single vessel occlusion, recovery of oxygen tension on the arteriolar wall was significantly faster than recovery of arteriolar diameters (P < 0.001). Following whole pouch occlusion in low oxygen suffusion, diameters often remained at their peak value for some time after complete recovery of periarteriolar PO 2. Thus, our measurements of periarteriolar PO 2 provide direct evidence that hypoxia of the resistance vessels is not present following release of the occlusion. This finding supports the conclusions of other investigators 16 ' 17 who reported substantial hyperemic responses even when vessels were probably exposed to fully saturated arterial blood (and thus argued that the maintenance TABLE 4 Tissue PO 2 Changes in Response to 1 -Minute Occlusions of Single Arterioles or the Whole Vascular Bed Tissue POj changes (mm Hg) Whole pouch occlusion Suffusion Control Minimum Control Minimum 0% O 2 5% O 2 10% O, 11.0 ± ± 0.6 (19) [10] 14.8 ± ± 1.8* (12) [4] 22.8 ± ± 2.2*t (9) [6] 11.3 ± ±0.8 (9) [6] 17.9 ± ± 1.9* (8) [4] 20.8 ± ± 1.7't (11) [6] All values are expressed as mean ± SEM. Control values represent PO, immediately prior to occlusion, and minimum values refer to the lowest PO, reached during the occlusion. Parentheses indicate number of arterioles observed; brackets indicate the number of animals. ' P < 0.05 vs. same type of occlusion in 0% O, sufusion. t P < 0.05 vs. same type of occlusion in 5% O, suffusion.

5 550 CIRCULATION RESEARCH VOL. 41, No. 4, OCTOBER 1977 of reactive hyperemia was not due to the direct effects of oxygen deficiency on the vascular smooth muscle). While Olsson 18 has suggested that some time may elapse before the vascular smooth muscle cells recover from hypoxia (which could explain their relaxation in spite of an abundant oxygen supply in the postocclusion period), comparison of the changes in arteriolar diameter and periarteriolar PO 2 in response to single vessel occlusion and whole bed occlusion during 0% O 2 suffusion suggests that this is unlikely. The minimum periarteriolar PO 2 and the recovery rate of periarteriolar PO 2 are not greatly different for single vessel and whole pouch occlusion, but the half-time for recovery of arteriolar diameters is substantially longer following release of whole bed occlusion (Tables 2 and 3). Furthermore, there is good evidence that the arterioles of the cheek pouch are not particularly sensitive to changes in PO 2 which are restricted to the environment of the smooth muscle. 19 Some studies 20 ' 21 have suggested that vascular smooth muscle is sensitive to oxygen deficiency at a fairly high PO 2. However, oxidative phosphorylation of hamster mesenteric arterioles appears to be highly resistant to hypoxia, and limitation of oxygen consumption does not begin to appear until an environmental PO 2 of 2-5 mm Hg is reached. 22 Data from isolated vascular strips exhibit a size dependence for the critical PO 2 of arterial smooth muscle and, based on an average arteriolar wall thickness of 10 /j.m, 2 mm Hg has been estimated to be the PO 2 at which the vascular smooth muscle should relax. 23 These findings indicate that the vascular smooth muscle is not limited directly by oxygen availability during occlusion in the 5% O 2 and 10% O 2 suffusions, since periarteriolar PO 2 did not fall below the value of 2-5 mm Hg. Moreover, the minimum periarteriolar PO 2 during occlusion was close to or greater than the preocclusion value of tissue PO 2 during 0% O 2 and 5% O 2 suffusion. Thus, if dilation of the arterioles were the direct result of oxygen deficiency during 5% O 2 and 10% O 2 suffusion, the vascular smooth muscle would have to be more sensitive to oxygen limitation than the parenchymal cells. During several occlusions in 0% O 2 suffusion, however, periarteriolar PO 2 approached 0 mm Hg. In such cases, arteriolar dilation could be at least partially due to the direct effect of oxygen deficiency on the vascular smooth muscle. The direct relationship between the minimum tissue PO 2 during occlusion and the PO 2 of the suffusion solution demonstrates that oxygen supply to the parenchymal cells can be increased by raising the oxygen content of the suffusion solution. Oxygen supply from the suffusate would undoubtedly become even more important for parenchymal cell function when blood flow is interrupted by occlusion. The increased oxygen supply to the tissue would then presumably result in decreased production of vasoactive metabolites during occlusion and, in turn, decreased arteriolar dilation. The decreased vasodilation we observed as the PO 2 of the suffusion solution was increased is consistent with this hypothesis. Our direct measurements of tissue PO 2 also confirm that hypoxia is not present in most of the tissue following release of the occlusion. The present investigation thus indicates that the maintenance of the hyperemic response is not due to the persistence of tissue hypoxia following release of the occlusion. This is consistent with the conclusions of both McNeill 16 and Eikens and Wilcken, 17 who observed reactive hyperemia under conditions in which tissue hypoxia should have been reduced or eliminated. These measurements of tissue PO 2 cannot completely eliminate the possibility of hypoxic foci, but the sites of measurement were chosen to lie in areas free of capillaries and, therefore, presumably represent the least oxygenated areas of the tissue. The present study suggests that oxygen-independent mechanisms can contribute as much as 60-70% to the dilation during occlusion (as estimated by comparison of the peak dilation during single vessel or whole bed occlusion in low and high oxygen suffusion). While the precise mechanisms causing arteriolar dilation during well oxygenated conditions cannot be determined from these experiments, several possibilities might be considered. 1. Arteriolar dilation during suffusion with the high oxygen solutions might reflect a myogenic response triggered by a decreased intravascular pressure during occlusion, However, the myogenic contribution to the steady state diameter changes during arteriolar occlusion in the hamster cheek pouch is apparently quite small Dilation during occlusion in high oxygen suffusion could be due to the action of aerobic metabolites. This hypothesis is supported by the results of several other studies 25 ' 28t 29 which have indicated that such metabolites might contribute to the hyperemic response. 3. The generation of vasodilator metabolites might be PO 2 -dependent, even at the high oxygen tensions we observed in the 5% O 2 and 10% O 2 solutions. These results are all consistent with an increasing body of evidence suggesting that regulation of oxygen availability by the circulation occurs at oxygen tensions substantially higher than the limiting PO 2 of the cytochrome oxidase system in isolated mitochondria. 30 ' 31 Acknowledgments We thank Elizabeth Staples and David Damon for their skillful technical assistance and Betty Haigh for an excellent job in typing the manuscript. References 1. Lewis T, Grant R: Observations upon reactive hyperaemia in man. Heart 12: , Fairchild HM, Ross J, Guyton AC: Failure of recovery from reactive hyperemia in the absence of oxygen. Am J Physiol 210: , Burton KS, Johnson PC: Reactive hyperemia in individual capillaries of skeletal muscle. Am J Physiol 223: , Gentry RM. Johnson PC: Reactive hyperemia in arterioles and capillaries of frog skeletal muscle following microocclusion. Circ Res 31: , Johnson PC, Burton KS, Henrich H, Henrich U: Effect of occlusion duration on reactive hyperemia in sartorius muscle capillaries. Am J Physiol 230: , Duling BR, Berne RM: Longitudinal gradients in periarteriolar oxygen tension; a possible mechanism for the participation of oxygen in local regulation of blood flow. Circ Res 27: , Duling BR: The preparation and use of the hamster cheek pouch for studies of the microcirculation. Microvasc Res 5: , Jacobson EG, Swan KG: Hydraulic occluder for chronic electromagnetic blood flow determination. J Appl Physiol 21: , Whalen WJ, Riley J, Nair P: A microelectrode for measuring intracellular PO,. J Appl Physiol 23: , Lombard JH, Duling BR: Relative contribution of passive and myogenic factors to diameter changes during single arteriole occlusion in

6 TWO-DIPOLE RANGING TECHNIQUE/M/rvis et al. 551 the hamster cheek pouch. Ore Res 41: , Whalen WJ, Nair P: Intracellular PO, and its regulation in resting skeletal muscle of the guinea pig. Circ Res 21: , Whalen WJ, Nair P: Skeletal muscle PO,; effect of inhaled and topically applied O, and CO,. Am J Physiol 218: , Coburn RF, Mayers LB: Myoglobin O, tension determined from measurements of carboxymyoglobin in skeletal muscle. Am J Physiol 220: 66-74, Duling BR: Microvascular responses to alterations in oxygen tension. Circ Res 31: , Whalen WJ, Nair P, Buerk D, Thuning CA: Tissue PO, in normal and denervated cat skeletal muscle. Am J Physiol 227: , McNeil! TA: Venous oxygen saturation and blood flow during reactive hyperemia in the human forearm. J Physiol (Lond) 134: , Eikens E, Wilcken DEL: Reactive hyperemia in the dog heart; effects of temporarily restricting arterial inflow and of coronary occlusions lasting one and two cardiac cycles. Circ Res 35: , Olsson RA: Myocardial reactive hyperemia. Circ Res 37: , Duling BR: Oxygen sensitivity of vascular smooth muscle. II. In vivo studies. Am J Physiol 227: 42-49, Detar R, Bohr DF: Oxygen and vascular smooth muscle contraction. Am J Physiol 214: , Carrier OC Jr, Walker JR, Guyton AC: Role of oxygen in autoregulation of blood Dow in isolated vessels. Am J Physiol 206: , Howard RO, Richardson DW, Smith MH, Patterson JL Jr: Oxygen consumption of arterioles and venules as studied in the Cartesian diver. Circ Res 16: , Pittman RN, Duling BR: Oxygen sensitivity of vascular smooth muscle. I. In vitro studies. Microvasc Res 6: , Dornhorst AC, Whalen RF: The blood flow in muscle following exercise and circulatory arrest; the influence of reduction in effective local blood pressure, of arterial hypoxia, and of adrenaline. Clin Sci 12: 33-40, Olsson RA: Kinetics of myocardial reactive hyperemia blood flow in the unanesthetized dog. Circ Res 14/15 (suppl I): 81-86, Patterson GC: The role of intravascular pressure in the causation of reactive hyperaemia in the human forearm. Clin Sci 15: 17-25, Hilton SM: Experiments on the post-contraction hyperaemia of skeletal muscle. J Physiol (Lond) 120: , Kontos HA, Patterson JL Jr: Carbon dioxide as a major factor in the production of reactive hyperaemia in the human forearm. Clin Sci 27: , Kontos HA, Mauck HP Jr, Patterson JL: Mechanism of reactive hyperemia in limbs of anesthetized dogs. Am J Physiol 209: , Rubio R, Wiedmeier VT, Berne RM: Relationship between coronary flow and adenosine production and release. J Mol Cell Cardiol 6: , Reivich M, Coburn R, Lahiri S, Chance B, (eds): Proceedings of the Symposium on Tissue Hypoxia and Ischemia, Philadelphia, August New York, Plenum, 1977 Detection and Localization of Multiple Epicardial Electrical Generators by a Two-Dipole Ranging Technique DAVID M. MIRVIS, FRANCIS W. KELLER, RAYMOND E. IDEKER, JOHN W. COX, JR., ROBERT F. DOWDIE, AND DAVID G. ZETTERGREN SUMMARY The ability of a numerical procedure to detect and to localize two experimentally induced, epicardial dipolar generators was tested in 24 isolated, perfused rabbit heart preparations suspended in an electrolyte-filled spherical tank. Electrocardiograms were recorded from 32 electrodes on the surface of the test chamber before and after placement of each of two epicardial burns. The second lesion was located either 180, 90, or 45 from the first. Signals were processed by iterative routines that computed the location of one or two independent dipoles that best reconstructed the observed surface potentials. The computed single dipole accounting for 99.68% of root mean square (RMS) surface potential recorded after the first bum was located 0.26 ± 0.10 cm from the centroid of the lesion. Potentials recorded after the second lesion were fit with two dipoles that accounted for ± 1.51% of RMS surface potentials and that were located 0.42 ± 0.26 cm and 0.57 ± 0.49 cm from the centers of the corresponding burn. Seventy-one percent of computed dipoles were located within the visible perimeter of the burn. Thus, two simultaneously active dipolar sources can be detected and accurately localized by rigorous study of the generated electrical field. CLINICAL electrocardiography strives to semiquantitatively define the physiological state of the heart from the electrical potentials it generates. The concept of an equivalent cardiac generator has been useful in this effort. An From the Section of Medical Physics, Department of Medicine, University of Tennessee Center for the Health Sciences, Memphis, Tennessee. Supported by Grants HL-01362, HL-09495, and HL from the National Heart, Lung and Blood Institute, National Institutes of Health, U.S. Public Health Service. Dr. Mirvis was supported in part by National Service Award HL from the National Heart, Lung and Blood Institute, National Institutes of Health. Dr. Ideker's current address is the Department of Pathology, Duke University, Durham, North Carolina. Address for reprints: David M. Mirvis, M.D., 951 Court Ave., Room 339M, Memphis, Tennessee Received November 30, 1976; accepted for publication March 25, equivalent cardiac generator may be defined 1 as a distribution of electrical sources in a specified volume conductor which generates potential distributions identical to, or "equivalent" to, those generated by the natural electrical generator, i.e., the heart. Many generator models have been proposed and tested in the search for a truly equivalent cardiac generator. The earliest and simplest generator was the fixed, single dipole model. Waller 2 idealized the relationships between the electromotive force of the heart and surface leads by considering the heart to be a lumped point source and point sink of current located within the cardiac region of the torso. Einthoven et al. 3 brought the source and sink pair close together to form an electrical doublet, or dipole.