I 119 231 unc‘SIs 5'3. 1', .;-..~. .0 -. Z‘vtmfzmlggmfi fitate Univeesfity This is to certify that the thesis entitled Investigation of Adenosine and Prostacyclin in Local Hypoxic and Hypercapnic Vasodilation in the Forelimb of the Dog. presented by Maureen T. Mulrenan has been accepted towards fulfillment of the requirements for M.S. Physiology Jegree in AM“ A» Mn Major professor Date W (9; “182. 0-7639 MSU LIBRARIES 5325-. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. INVESTIGATION OF ADENOSINE AND PROSTACYCLIN IN LOCAL HYPOXIC AND HYPERCAPNIC VASODILATION IN THE FORELIMB OF THE DOG By Maureen Therese Mulrenan A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Physiology 1982 ABSTRACT INVESTIGATION or ADENOSINE AND PROSTACYCLIN IN LOCAL HYPOXIC AND HYPERCAPNIC VASODILATION IN THE FORELIMB or THE DOG By Maureen Therese Mulrenan The role of adenosine and prostacyclin in local hypoxic and hypercapnic vasodilation was investigated in isolated, innervated canine forelimbs. Blood from the femoral artery was pumped at constant flow through an extracorporeal lung obtained from a second dog and then to the brachial artery of the forelimb. Gases ventilating the extracorporeal lung were varied at ten minute intervals: normoxia, 15-20% 02; mild hypoxia, 5% 02; severe hypoxia, 0% 02; hypercapnia. 15% C02. After gas tension alterations. blood samples were obtained from the brachial artery, brachial vein, and cephalic vein for adenosine and prostacyclin analysis. Hypoxia and hypercapnia significantly (p<0.05 or p<0.01) decreased forelimb resistance and perfusion pressure, yet plasma concentrations of adenosine and prostacyclin in the vessels draining the forelimb did not increase. Forelimb levels of adenosine and prostacyclin averaged 0.10uMolar and 1.9ng/ml plasma, respectively. These findings suggest that adenosine and prostacyclin are not involved in the vasodilation associated with local hypoxia or hypercapnia. DEDICATION This thesis is dedicated to Dr. Jerry Benjamin Scott whose untimely death did not allow him to see its completion. I am grateful for the opportunity I had to work with such a scholar. But more importantly I am thankful for the sense of honesty and justice he instilled in me with respect to scientific research. His work, words and life touched so many, and he will always be remembered in a special way by those who knew him. ii ACKNOWLEDGMENTS I wish to thank Dr. Scott w. Walsh for his encouragement and willingness to assist when I needed help so much. I can not fully express my gratitude for all that he has done in helping to complete my Masters training. Additionally, I wish to thank Drs. Thomas E. Emerson, N. Edward Robinson, and Harvey V. Sparks for their assistance during my graduate training. Finally, I want to express my appreciation to those who helped on a day-to-day basis with my prepartions and analyses: Dr. John P. Manfredi, Mr. Gregory D. Romig, Mr. Joel L. Silver for help with the adenosine analysis; Mr. Donald L. Anderson, Mr. Ronald J. Korthius, Dr. Neil C. Olson, Ms. Cindy A. Delonjay, and Mr. William V. Stoffs, for assistance during blood gas sampling and surgery; Ms. Allison Pankratz for my thesis typing; and Ms. Amylou Davis for her friendship and clerical assistance. iii TABLE OF CONTENTS Page LIST OF FIGURES...OOOOOOOOOOOOOO.OOOOOOOOOOOOOOOOOOOOOOOO Vi I. LITERATURE REVIEWOOOOOOOOOOOOOOIOOOOOOOOOOOOOOOO... 1 A. IntrOdUCtionIO00......OOOOOOOOOOOOOOOOOOOOO0.0. 1 B. Factors Affecting Vascular Smooth Muscle Tone During Hypoxia and/or Hypercapnia.............. 2 1. DireCt Effect or oxygen. 0 O I O O O O O O O O O O O O O O O O 2 2. PotaSSj-umOOOOOOOOOOOOOOOOOOOOOOOOOIDOOIOO0. 6 3O 05m01arity000000000OOOOOOOOOOOOOOOOOOOOOOO. 7 u. Hydrogen Ion and Carbon Dioxide............ 8 5. Adenine Nucleotides and Adenosine.......... 9 6. ProstaglandinSOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 11 C. Introduction to Thesis Study................... 13 II. MATERIALS AND METHODS. O O O O O O O O O O O O I O O O O O O I O O O I O O O O O 1S A. Isolated Forelimb and the Extracorporeal Lung.. 15 1. PreparationOOOOOOOOOOOOOOOOOOO00.0.0.0...O. 15 2. Experimental Protocols..................... 19 a. Hypoxia and Hypercapnia................ 19 b. Forelimb Blood Flow Determinations..... 20 B. AdenosineOOOOOOOOOO0.0.0.0...OOOOOOOOOOOOOOOOOO 21 1. Sample Collection and Preparation.......... 21 2. Analysis.OOOIOOOOOOOOOOOOOOIOOOOOOIOOOOOOO. 22 C. Prostacyclin (6-keto-prostaglandin F1 alpha)... 2n 1. sample CalleCtion. O O O O O O I O I O O O O O O 0 O O I O O O O O O 2” 2. RadiOimmunoassay. O O O I O O O O O O O I O O O O O O O O O O O O O O 2” 3. Infusions of Prostacyclin.................. 25 D. statistical AnalySis. O O O O O O O O O O O O O O O O O O O O O O O O O O 26 iv TABLE OF CONTENTS--continued Page III. RESULTS.00......OOOCOOCOCOOOOOCCOOOOO ..... 000...... 27 A. Perfusion Pressure, Resistance and Blood Flow.. 27 B. AdenOSineOOOOOOOO00......OOOOOOOOOOOOCOOOOOOOOO 27 C. Prostacyclin (6—keto-prostaglandin F1 alpha)... 3a Iv. DISCUSSION.00.00.000.000..0.OOOOOOOOOOOOOOOOOOOOOOO ”2 V. SUMMARY AND CONCLUSIONS ............ . . . . . . . . . . . . . . . . 117 LIST OF REFERENCES. 0 O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 0 ..... O “8 APPENDIXOOOO ..... 00......OOOOOOOOOOOOOOIO0.00... ......... S” FIGURE 1. LIST OF FIGURES Diagram of the isolated forelimb preparation with an extracorporeal lung .0.0.0.0.0000...OIOOOOOOOOOOOOOOOO Perfusion pressure and resistance in the forelimb during normoxia, mild hypoxia, severe hypoxia, and hypercapnia....OOOIOOOOOOOOIOIOO0.0.0.0...00.000.00.00 Blood flow through the brachial and cephalic veins during control and severe hypoxia..................... Concentrations of adenosine in plasma water during normoxia, mild hypoxia, severe hypoxia and hypercapnia....oOIOOOOOOOOIOOO0.000000000000000000000. Concentrations of 6-keto—prostaglandin F alpha during normoxia, mild hypoxia, severe hypoxia a d hypercapnia....OOOOOOO0.00.00.00.00...IOOOOIOOIOOOIOO. 6-keto—prostaglandin F alpha measurements during prostacyclin infusion into the isolated forelimb...... Systemic blood pressure, forelimb perfusion pressure, and resistance during infusion of prostacyclin into the forelimbOOIOOOOOOO0.0.00....OOOOIOOOOOOOOOOOOOOOO. vi Page 16 28 3O 32 35 37 40 I. LITERATURE REVIEW A. Introduction Since the late 1800's there has been a growing concern and interest in understanding the mechanism by which blood vessel diameter changes. Vessel diameter is an important determinant of resistance which is important in hypertension and vascular disease. Luigi Severini first proposed that both oxygen and carbon dioxide can directly alter vessel diameter (62). This proposal, however, was quickly opposed by W.H. Gaskell who hypothesized that metabolites surrounding the vessel were the main determinants of its contractile state (23). The subject was readdressed occassionally during the following 80 years, but it was not until 1964 that interest in this area surged again. At this time it was reported that isolated arterial segments challenged with various blood oxygen tensions below lOOmmHg contracted less than normoxic strips (8). Investigators then began looking for a specific mechanism to determined the mechanism of the decrease in vessel wall activity with lowered P02. In addition to oxygen and carbon dioxide, other metabolites have also been proprosed as mediators of the vasoactive state of the blood vessel wall. The following is a brief summary of the current knowledge on various factors known to affect blood vessel diameter. Because there is little information available on skeletal muscle preparations, this review is limited primarily to data obtained from isolated arterial strips. B. Metabolic Factors in the Local Control .22 Blood Flow in Skeletal Muscle 1. Direct Effect g§_0xygen This review will begin with the data supporting the direct action of oxygen ("1 the blood vessel wall. The vascular wall needs a constant supply of oxygen to perform work. If the supply is decreased, it seems logical that the amount of work, i.e. contraction, that the vessel could perform would decrease. This has lead many investigators to the hypothesis that vasodilation results directly from a vessel wall oxygen debt. In 1967 Daugherty et. al. (11) noted that a decrease in perfusion pressure occurred when the P0 of arterial blood perfusing an isolated 2 skeletal muscle of the dog was lowered by use of an extracorporeal lung. The critical perfusate P0 was approximately uOmmHg. Perfusion pressure 2 remained steady as the P0 was decreased from 100mml-lg to llOmmHg but 2 beyond this a concomitant drop in perfusion pressure was seen with each successive lowering of the oxygen supply to the muscle. Ross et. al. (5“) studied central effects involved with the vasodilation. In a series of experiments, one group of dogs had their spinal cords severed and the spinal cords of the other group remained intact. An isolated hind leg preparation was used. Initally both groups received blood of normal oxygen saturation, 100 percent, for the control period. Then blood from the vena cava was perfused through the hind leg. In both groups of dogs, a decrease in perfusion pressure occurred with this hypoxia and blood flow increased over three times the resting level. Ross et. al. concluded that the brain was not involved in this autoregulatory mechanism. A question arose concerning the use of systemic arterial blood PO2 as an index of the oxygen environment of the vascular smooth muscle cells. Duling et. al. (17) addressed this question using the intact hamster cheek pouch and the cremaster muscle of hamsters and rats. They measured arteriolar and tissue P02 and feund that the arteriolar wall PO2 gradient was small, an average of 1.MmmHg different from inside to outside. Thus, the use of arterial blood P02 as an index of vessel wall PO2 seemed appropriate because there did not appear to be a substantial barrier for oxygen diffusion across the vessel wall. Furthermore, the tissue P0 was always a few mmHg lower than either the vessel wall 2 measurement or the arterial P02. Thus, arterial P02 is a better index of vessel wall PO2 than is tissue P02. Once this was accepted, Duling (l6) devised a way to separate the direct action of oxygen on the vessel from the actions of metabolic agents. By superfusing a small vascular portion of a hamster cheek pouch with a micropipet while supplying a different perfusion solution to the tissue he could measure arteriolar diameter under various conditions. With the assistance of PO electrodes placed perivascularly 2 and on the apposite side of the vessel from the superfusion area, PO2 was measured and correlated with changes in vascular diameter. Several experimental gas changes were performed and in all approaches changes in the perfusate gas tensions to the tissue had a greater influence on vessel diameter than did changes in the P02 directly around the vessel. Duling concluded that a vasoactive metabolite may be released by the tissue to control vessel caliber. Because the perivascular PO2 electrode was placed on the opposite side of the smooth muscle from the lumen, and the blood that perfused the lumen was high in oxygen content, it was not possible to know the oxygen content of the vascular smooth muscle. Numerous studies have been performed on isolated vascular strips contained in a bath. Tissue baths maintain the strips at a constant body temperature in a physiological salt solution. These in £19.19. preparations allow central and extravascular mechanisms to be eliminated, so that local control of vascular contraction can be studied. Smith et. al. (63) superfused arterial strips from various animals (cat, guinea pig, rat, pig) with a physiological salt solution with various oxygen contents. Vascular reactivity correlated directly with the oxygen content of the superfusate. This effect was unaltered by' hyoscine, phenoxybenzamine, hexamethonium, bromolysergic acid diethylamine or mepyramine indicating that the parasympathetic and sympathetic systems were not involved, nor serotonin or histamine. The direct correlation of vascular reactivity to oxygen has been corroborated by numerous other investigators using different types of arterial segments and species (14, 20, 25, 51, 65). In some studies it was found that vessels responded differently depending on vessel wall thickness. Arterial strips with thick vascular walls had decreased contractile responsiveness at higher oxygen concentrations than thinner vessels, presumably because the core of the thicker vessels became hypoxic sooner due to a larger diffusion gradient (51). In both thick and thin vessels, however, decreased P02 resulted in a decreased contractile state. Two hypotheses have been proposed to explain the relationship between oxygen and vessel wall diameter. The first states that relaxation occurs because of the decreased ability of the vessel to produce energy (oxidative phosphorylation). The second deals with the liberation of some unknown vasoactive substance. It is easy to understand how a decreased oxygen supply would interfere with tension development by eliminating a necessary component for oxidative phosphorylation. Fay et. al. (20) isolated ductus arteriosus strips from newborn guinea pigs and placed them in a muscle bath. Inhibitors of oxidative phosphorylation only inhibited the contractile response to oxygen and had little effect on acetylcholine induced contraction. Carbon monoxide also inhibited the oxygen response. This latter inhibition could be reversed by light, most likely due to photodissociation of the cytochrome a -C0 complex. These 3 data suggest that oxygen stimulates contraction probably by enhancing the rate of oxidative phosphorylation. Namm et. al. (A7) studied rabbit aortic strips in an incubation bath. They found that the oxygen concentration in the bath was poorly correlated with ATP levels liberated into the bath by the aortic strips. They could not, however, dismiss the energy-available hypothesis because of possible compartmentalization of ATP. These investigators theorized that ATP pools may be in smooth muscle and it is these small pools that provide the energy for tension development. If this is so, the correlation between ATP and contraction may not be possible to elucidate. Further, the decreased creatine phosphate that they observed may be from the equilibration with this extramitochondrial ATP pool. And furthermore, the nonmuscular cells producing ATP may be responsible for the decrease of ATP that is actually occurring within the contractile compartments. In addition, Gellai et. al. (25) noted that the decreased responsiveness of strips bathed in an hypoxic medium does not wane with time. Using arterial strips, they demonstrated a 35-40 percent decrease from control tension with a PO2 change of 100 to 10mmHg. This response was tested for one hour and it was sustained over the entire period. Chang et. al. (9) measured the PC at the surface of the arterial 2 wall of rabbit thoracic aortas, deep femoral arteries and skeletal muscle arteries using a muscle bath preparation and oxygen-sensitive microelectrodes. From the direct PO2 measurements, the oxygen tension within the arterial wall was estimated as the oxygen concentration in the muscle bath was reduced. A decreased contraction of the vessel was seen with each reduction. below ‘the estimated arterial wall P02 of 50mmHg. ‘This hypoxic PO estimation of 50mmHg was the first data to 2 show that vasodilation occurred at physiological oxygen tensions within the muscle strip. In summary, these data elucidate the role oxygen plays in directly influencing vessel tone. Decreased oxygen causes relaxation of vessels .lfl.l£££2v as well as, vasodilation in xixg, Decreased contractile state is correlated with decreased oxygen supply. The data indicate that the direct action of oxygen plays an important role in the vasoactive state of blood vessels in the peripheral circulation. 2. Potassium In 1938, Katz et. al. (36), using Langendorff preparations of dog hearts showed that potassium can produce brief dilation. Increasing the concentration of potassium in the blood to the coronaries from 220mgm. percent to 279mgm. percent dilated the coronary bed, but within a few minutes there was a return to control tone and blood flow. Decreasing the potassium concentration to the coronary circulation of the dog, in 111g, caused vasoconstriction (29). When potassium was infused into the arterial blood supply of the forelimb or hindlimb of the dog vasodilation occurred (7). Brace (7) characterized this response by noting that there was a rapid decrease in vessel tone, but the vasodilation waned within the following few minutes. Many types of vessel strips relax when the concentration of potassium is increased (25, 67). Gellai et. al. (25) showed that inducing relaxation of rabbit coronary and skeletal muscle strips in a bath by addition of potassium (bath concentration = umM) only produced a transient relaxation. This response was a 40 percent decrease in contraction from control and the strips recovered to 100 percent of their control tension within five to six minutes. Similarly, Toda (67) induced transient relaxation by increasing the bath concentration of potassium to 5mM. The relaxation was unaffected by tetrodotoxin or propranolol which suggests that the release (of neurotransmitters and beta adrenergic mechanisms were not involved. A potassium-free medium can cause contraction of strips, but relaxation can then be induced with addition of potassium (as low as 0.1mM) (6). These findings indicate that potassiwm does not have a role in a sustained vasodilation which would rule out its involvement in hypoxic or hypercapnic vasodilation. 3. Osmolarity Coronary and deep femoral artery strips from New Zealand white rabbits have been studied in muscle bath preparations to determine the vasodilation induced by hyperosmotic solutions (25). A bath concentration of 30 milliosmoles/L induced a 20410 percent relaxation from control tension. However, within 15-60 minutes the strips recovered to control contractile tension. Addition of potassium (bath concentration = llmM) to this high osmolar solution relaxed strips to 85-95 percent of control tension but recovery was 100 percent complete within 10-15 minutes. Although the additive effect of potassium produced a much larger relaxation than the hyperosmotic medium alone, recovery was still complete within a short time. Other studies (24, 60, 64) have also concluded that sustained vasodilation does not result from a hyperosmolar environment. Thus, osmolarity does not seem to have a role in long term vasodilation as is induced by hypoxia or hypercapnia. n. Hydrogen Ion and Carbon Dioxide In 1962, Molnar et. al. (flu) infused isoosmotic acids (hydrochloric, nitric, lactic, pyruvic, acetic and citric acids) intra-arterially into the forelimb of the dog to evaluate the resistance changes occurring with decreased pH. Resistance decreased during all infusions, as did arterial and venous pH (mean of 7.36 to 6.72, and 7.39 to 6.85, respectively). They concluded that hydrogen ion was locally vasoactive. In 1968, Kontos et. al. (39. HO) examined the vasodilation associated with local hypercapnic acidosis and breathing 7 percent CO2 in the human. Four acid phosphate buffer solutions were infused intra-arterially at successively increasing rates and constant pressure into the intact forearm (H0). Venous pH was lowered from 7.3” to 7.2a with a concomitant increase in venous P002 of NZ to 52mmHg. Blood flow to the forelimb increased from A to 7mfl/min/100g tissue. Hypercapnia induced systemically by 7 percent CO breathing also decreased vascular 2 resistance in the intact forearm. Phenoxybenzamine and propranolol were given to block sympathetic effects (39). When the subjects breathed the hypercapnic gases and received a sodium bicarbonate infusion to block the acidosis, no change in vascular resistance was noted. Therefore, the authors concluded that the increase in blood PC02, not the decrease in pH induced by hypercapnia. was responsible for the vasodilation. However, Rooke et. al. (53) placed isolated canine coronary arteries and saphenous veins in a tissue bath while altering pH or PC02 and noted that extracellular pH influenced the vessel tension, independent of PCO Vessels relaxed when the bath pH was lowered from 7.11 and 2. constricted at pH values above this. Alterations in PC02 from 20-56mmHg were not correlated with vessel activity. The data combined show that the mechanism of hypercapnic vasodilation is still undetermined. Some investigators hypothesize that the increase in the carbon dioxide tension directly effects vessel tone, while others suggest that pH alterations correlate with vessel reactivity. 5. Adenine Nucleotides and Adenosine The vasodilatory action of the adenine nucleotides were first noted in the general arterial and coronary vasculature in 1929 (15). Since then, adenosine triphosphate (ATP), adenosine diphosphate (ADP). adenosine monophosphate (AMP) and adenosine have all been recognized as potent vasodilators in most vascular beds (2, 10, 12, 66) including the canine forelimb (22, 28). Rubio et. al. (55) performed histochemical studies. on. skeletal muscle from rats and guinea pigs in order to characterize the pathway of adenine nucleotide degradation. Activity of 5' nucleotidase, the enzyme 10 that produces adenosine from nucleotides, was found to be localized in and near the endothelium of the blood vessels. Other areas of the skeletal muscle degrade nucleotides by the inosinic acid (IMP) pathway. Herlihy et. al. (30) demonstrated that pig carotid strips which had been contracted with norepinephrine and potassium were subsequently relaxed by exogenous adenosine (3 X 10-6M). Further, pig carotid artery strips incubated during normoxia (95102:5%C02) and anoxia (957N2:5%COZ) demonstrated a five-fold increase in hypoxanthine and a two-fold increase in inosine during anoxia (68). Tissue ATP levels on the other hand decreased with anoxia. Because adenosine added to the medium was rapidly deaminated to inosine, the authors concluded that adenosine was formed during anoxia by the arterial strips and was rapidly degraded it to inosine. Hypoxia and hypercapnia were induced in the isolated forelimb of the dog by Kienitz et. al. (37) and the changes in perfusion pressure were observed. TheOphylline (2.7umg/min), a blocker of the adenosine receptor, was infused and hypoxia and hypercapnia were again induced. Vasodilation was not inhibited by 'theophylline. The investigators concluded that adenosine did not appear to be involved in the hypoxic or hypercapnic vasodilation. However, due to the vasodilatory action of theophylline, norepinephrine was needed to raise perfusion pressure to pre-infusion levels. The data on adenosine in local hypoxia and hypercapnia are inconclusive and contradictory. Some investigators conclude that there is a role for adenosine in local vasodilation, and others that adenosine is not involved. The release of adenosine during local hypoxia and hypercapnia in skeletal muscle has not been investigated. 11 6. Prostaglandins Prostaglandins associated with anoxia were first studied in the coronary vascular bed by Block et. al. (3. 4). The coronary beds of isolated rabbit hearts were perfused with a Krebs solution containing 95102:5$C02. A prostaglandin, postulated as E2, increased and its increase could be blocked by indomethacin. However, the coronary vasodilation seen with anoxia was not blocked by indomethacin. Thus, the authors concluded that the prostaglandin released during anoxia did not have a vasodilator role. Criticism arose concerning these studies because the cyclooxygenase enzyme that catalyzes the arachidonic acid cascade needs oxygen to produce prostaglandins. The mechanism behind the increased prostaglandin production during anoxia, therefore, remained a mystery. Wennmalm et. a1. (70) also used isolated rabbit hearts, but induced hypoxia, 510 rather than anoxia. They found that there was a 2. decreased production of a prostaglandin E-like compound during hypoxia, but production of this compound increased after hypoxia was ended. The investigators concluded that prostaglandins may be involved in reactive hyperemia, but not under conditions of low oxygen. Two years later in 1976, Kalsner (3A) placed isolated bovine coronary artery strips in a muscle bath and induced hypoxia. He found that as the arterial strips were challenged with decreasing bath oxygen tensions (515 to 38mmHg), the release of a prostaglandin E-like substance increased. Contrary to Wennmalm et. al. (70), Kalsner concluded that vascular strips that relaxed when challenged with low oxygen produced a prostaglandin that was correlated with the decreased contractile state of the strip. 12 Kalsner followed his previous study with another (33) in which the oxygen content of the bath medium surrounding the isolated coronary arterial strips was lowered to 9mmHg. At this low oxygen concentration release of the prostaglandin E-like compound stopped and the strips contracted. Kalsner proposed that vessel strips needed a supply of oxygen to produce a vasodilator prostaglandin, otherwise, contraction occurred. Detar (13) challenged the work of Kalsner and placed rabbit skeletal and cardiac muscle arteries in a tissue bath and induced hypoxia stepwise from a P0 of 60 to 10mmHg. Contraction of the arterial strips 2 was depressed with hypoxia but this depressed activity was not affected by indomethacin (10-5 M=bath concentration). No data was provided in the paper and sample size was only two. Detar, nevertheless, concluded that hypoxic vasodilation was due to the direct effect of oxygen and did not involve prostaglandins. Hypoxia and hypercapnia were also studied in the pial arterioles of the intact cat brain (69). Increasing the perfusing carbon dioxide tension or decreasing the oxygen tension in the artery caused vasodilation. This response was not significantly altered when two cyclooxygenase inhibitors (indomethacin or ARR-5850) were infused. The blockers did, however, substantially inhibit the vasodilator response seen with topical application of“ arachidonate (100-200ug/ml). The authors concluded that endogenous prostaglandins are not involved in the hypoxic or hypercapnic responses in the pial microcirculation. However, Pickard et. al. (99) obtained data contrary to this. They altered the carbon dioxide tension from no to 60mmHg locally to the brain of the baboon and saw an increase in cerebral blood flow from 57 to 13 110m1/100g/min. Yet, after administration of indomethacin (0.014 to 0.2mg/kg/min) through a lingual artery cather, blood flow' did not increase with hypercapnia. However, resting blood flow was lower after pre-treatment with indomethacin, NOml/lOOg/min, which constricted the vascular bed prior to hypercapnia. The vasodilation associated with hypoxia or hypercapnia was studied, 1 vivo, by Kienitz et. al. (37) in the isolated forelimb of the dog. Vasodilation was observed with both hypoxia and hypercapnia and infusion of indomethacin (1.72mg/min) did not 'block the vasodilation. The results failed to provide positive support for the involvement. of prostaglandins in the vasodilation with hypoxia or hypercapnia, however, prostaglandin levels were not measured. The literature implicating a role for prostaglandins in local hypoxia or hypercapnia is controversial and not conclusive. Some prostaglandins do produce vasodilation when infused or administered but they do not appear to be necessary for dilation of blood vessels during hypoxia or hypercapnia. Most of the studies in the literature in this area have used isolated vascular strips, and therefore, the.ig.gigg data is very limited. Also the vasodilation associated with local induction of hypoxia and hypercapnia in skeletal muscler needs further investigation. 7. Introduction £2.IEEEEEH§EEQX I attempted to uncover more information about the possible role of adenosine and prostacyclin in the vascular response to local hypoxia and hypercapnia. An isolated, innervated forelimb preparation was used similar to that was used by Kienitz et. al. (37) and the forelimb arterial and venous plasma concentrations of adenosine and prostacyclin 14 during local hypoxia or hypercapnia were measured. Femoral arterial plasma samples were also drawn to compare our prostacyclin concentrations with the results of other investigators (111). Since prostacyclin is the major product synthesized via the cyclooxygenase enzyme from arachidonic acid in the vessel wall (32, 145), we assumed that prostacyclin would increase in the plasma if it was involved in the vasodilation induced by hypoxia or hypercapnia. Also, adenosine is a very potent vasodilator ighxixg and ignxitrg, and the vasculature of the skeletal muscle can produce adenosine. We assumed that if adenosine was involved in local hypoxia or hypercapnia, it could be measured in the venous plasma of the isolated forelimb. II. MATERIALS AND METHODS A. The Isolated Forelimb and the Extracorporeal Lung 1. Preparation Eighteen mongrel dogs (15-33 kg) of both sexes were anesthetized with sodium pentobarbital (35mg/kg.,iv., Abbot Labortories, North Chicago, IL). Supplemental anesthesia was administered as needed during experimental procedures. Each dog was intubated and the endotracheal tube was connected to a constant volume respirator (Harvard Apparatus Company, Model 613, Millis, MA.) that was supplied with room air and supplemented with 100 percent oxygen to achieve an arterial blood gas tension of 100 mmHg. An acid-base analyzer (PHM 72Mk2, Radiometer Copenhagen, Capenhagen, Denmark) was used to measure P0 P00 and pH. 2' 2 The right forelimb of the dog was surgically isolated, except for the major nerves, the brachial artery, and the brachial and cephalic veins (see Figure l). The right femoral artery and the left femoral artery and vein were exposed. The left femoral artery was cannulated with P.E. 290 tubing (Intramedic TUbing, Clay Adams, Parsippany, NJ) to obtain arterial blood samples and to continuously monitor systemic arterial blood pressure. Blood pressure was measured using a pressure transducer (Statham Laboratory, Model P23Gb, Hato Rey, Puerto Rico) and a Hewlett-Packard eight channel direct writing recorder calibrated daily againsted a mercury manometer (model 7796A, Boston, MA). The left femoral vein was cannulated for drug infusions. The median cubital 15 16 Figure _1__ Diagram of the isolated forelimb preparation with an extracorporeal lung. FA = femoral artery BA : brachial artery BV = brachial vein CV = cephalic vein Blood was pumped from the femoral artery to the extracorporeal lung and then to the isolated forelimb. The blood was drained via the intact brachial and cephalic veins. RESPIRATOR F LOW CONTROL l now ECONTROL PULMONARY VENOUS PRESSURE BRACHIAL ARTERY PRESSURE 4—- CEPHALIC VEIN PRESSURE 4---' BRACHIAL VEIN PRESSURE 4—“ \f/m ”5. / ‘ET ,‘I H!’\ RESPIRATOR FE MORAL ARTERY I ARTERIAL PRESSURE‘ (RECORDER) Figure 1 18 vein, which is the major anastomotic connection between the brachial and cephalic veins, was ligated and a cannula was inserted into each forelimb vein via the ligated vessel. These cannulae were used to measure brachial and cephalic venous pressures and to sample venous blood from the forelimb muscle and skin, respectively. Venous pressures were recorded as described above for the left femoral artery. An extracorporeal lung (ECL) was obtained from a second anesthetized dog (10-l2kg) that was injected intravenously with sodium heparin (10,000 USP units, Elkins-Sinn Inc., Cherry Hill, NJ). After ten minutes, 300-500cc of blood was taken to be used later for priming the ECL. A left thoracotomy was performed at the fifth intercostal space. The inferior vena cava was ligated and transected. The pericardial sac was opened and a ligature placed around the pulmonary artery for easy indentification later. Then the heart/lung unit was excised. The lower half of the left heart, as well as the entire portion of the right heart below the pulmonary arterial ligature was removed. The lung unit was rinsed with isotonic saline to remove residual blood. Meanwhile, a reservoir containing the heparinized blood obtained from the donor dog was used to prime a Masterflex blood pump (Cole-Parmer Instrument Co., Chicago, IL) and tubing for the pulmonary arterial cannula (P.E. 380). Sodium heparin (10,000 USP units) was administered to the recipient dog. A 1 % inch diameter tubing was inserted into the left atrium of the excised heart and positioned to receive pulmonary venous blood. The trachea from the excised lungs was intubated and ventilated with a second respirator. The blood primed pulmonary arterial cannula was inserted into the pulmonary artery. Blood was pumped from the beaker reservoir, through the pulmonary artery, the lung lobes, the pulmonary 19 vein and back to the beaker reservoir via a Holter roller pump (Extracorporeal Medical Specialties Inc., King of Prussia, PA). The right femoral artery of the recipient dog was cannulated (P.E. 2110) to supply blood to the pulmonary artery of the ECL unit. The venous cannula from the ECL was connected to the brachial artery. In this way, blood was pumped from the femoral artery of the recipient dog, through the ECL, the isolated forelimb and drained via the brachial and cephalic veins of the forelimb (Figure 1). Pulmonary arterial and venous pressures were continously monitored and recorded as described above. The pump supplying the ECL was regulated by a Leeds-Northrop controller (North Wales, PA) that adjusted flow to maintain a venous pressure of 2-6mmHg. Thus, capillary hydrostatic pressure in the ECL was regulated to avoid damage to the aveoli and to minimize pulmonary edema. Interposing the ECL between the recipient dog and the isolated forelimb made it possible to change gas tensions of the blood supplying the forelimb without altering systemic blood gas tensions. Before any experimental manipulations were performed, blood flow to the forelimb was adjusted until perfusion pressure approximated mean arterial blood pressure and remained constant throughout the study. The blood pressure, the perfusion pressure and the blood gases of the ECL were allowed to equilibrate for 20-30 minutes. After equilibration, systemic arterial blood gases were as follows: P02=123:6.5, Pcoz=3711.1, pN=7.37io.o1 (mean 1 SEM, n=18). 2. Experimental Protocols a. Hypoxia and Hypercapnia Changes in oxygen and carbon dioxide tensions were made to evaluate their vasodilator effects in the isolated forelimb. At ten minute 20 intervals, gas tensions to the forelimb were randomly altered by changing the gas mixture supplying the ECL. Two levels of hypoxia, one level of hypercapnia, and two control periods were performed in each eXperiment. The gas mixtures used were as follows: normoxia, 15-20% 02:51 002:75-801 N2; mild hypoxia, 5% 02:51 C02:90% N2: severe hypoxia, 0% 02:51 C02:95$ N2: hypercapnia, 15$ 02:l5$ C02:70$ N2. These experimental manipulations resulted in a brachial arterial blood P0 f 20 u9i2.5mmHg for mild hypoxia, a P0 of 21:1.7mmHg for severe hypoxia, and 2 a PC02 of 93:6.3mmHg for hypercapnia (in this latter case, pH decreased from 7.35:0.01 to 7.08:0.02, mean I SEM). During normoxia intervals P02 and PCO2 were adjusted to 100mmHg and uOmmHg, respectively, with a pH of 7.“. Each experiment began with a ten minute control period, followed by one or two experimental alterations, a control period, and the additional experimental alteration(s). Blood was taken at the end of each ten minute interval from: 1) femoral artery (systemic arterial blood); 2) brachial artery (blood coming from the ECL); 3) brachial vein (forelimb muscle); N) cephalic vein (forelimb skin). A steady state was usually attained during each ten minute interval with respect to perfusion pressure, systemic blood pressure, P02, PC02 and pH; however, in some cases up to 13 minutes were required. b. Forelimb Blood Flow Determination Pump flow (forelimb inflow) was measured at the end of each experiment with a graduated cylinder and a stop watch in all animals. Before sacrifice in six animals, one of the forelimb veins was cannulated to provide direct collection of venous blood. Venous outflow was measured using a graduated cylinder and stop watch at the end of a 21 ten minute control period, and after ten minutes of severe hypoxia. Total forelimb inflow minus the directly measured venous outflow gave the flow rate of the other vein. Any redistribution of blood flow between the brachial and cephalic veins could thus be determined. This procedure was completed during severe hypoxia and since no redistribution of blood occured, mild hypoxia was not tested. Hypercapnia was not studied because a previous study demonstrated no redistribution of forelimb blood flow during hypercapnia (52). B- W 1. Sample Collection and Preparation Blood was collected to determine whether adenosine was involved with the vasodilation induced by hypoxia and/or hypercapnia. Blood samples (approximately 2m1) were placed, within 30 seconds, in precooled, preweighed tubes containing 250ul of 3uM «erythro-9(2-hydroxy-3nonyl) adenine (EHNA, Burroughs Welcome, Research Triangle Park, NC), 0.26uM dipyridamole (Boeringher Ingleheim, Ridgefield, CT), and five percent methanol in isotonic saline. EHNA and dipyridamole were added to block the breakdown of adenosine and the uptake of adenosine by red blood cells, respectively, so that adenosine concentrations measured in the plasma would be indicative of forelimb adenosine production or uptake. Samples were then centrifuged (IEC Clinical Centrifuge, Needham Hts, MA) at 2800rpm (1360xg) and “ac. Four quality control tubes were included in each experiment, that is, a "spiked" tube containing the above collecting solution plus 20-25ul of adenosine (approximately 0.3nmole). and three tubes containing five or 20u1 (10 or 1.7mg/ml, respectively) of Type I 22 adenosine deaminase (Sigma, St. Louis, MO) without EHNA. The adenosine deaminase tubes were used to determine the purity of the unknown samples during analysis on the high pressure liquid chromatograph (HPLC). An adenosine deaminase sample was taken with each experimental alteration and analyzed along with its paired unknown sample taken at the same time and sampling site. Adenosine deaminase samples were set aside at room temperature for ten minutes before centrifugation to allow adenosine deaminase to break down adenosine to inosine and hypoxanthine. Otherwise, they were processed as the other samples. One ml of plasma from each sample was pipetted and placed into 250ul of a 35 percent perchloric acid solution, mixed and centrifuged (Sorvall model RC2-B, Newton, CT) at “CC and 17.500rpm (32,000xg) for 15 minutes. The supernatant was decanted and 900ul transferred to another glass tube. These samples were then neutralized with llOul of K (1.0g/ml) 2CO3 solution to a pH of 6.5-7.5, mixed and centrifuged (IEC Centra-7R, International Equipment Co., Needham Hts, MA) at u°c and 2800rpm (1360xg) for ten minutes. The supernatant was decanted into a new tube and frozen (-20°C) until the time of assay. The collection tubes were later weighed to calculate the plasma volume that had been collected. 2. Analysis Adenosine was measured by high pressure liquid chromatography (HPLC) which achieves high selectivity of a compound. An extensive discussion of the procedure and validation of the adenosine assay used can be found in the doctoral thesis by John Paul Manfredi (”2). Briefly, samples (100ul) were injected into a reversed-phase column (either uBondapak C18, Waters Associates, Milford, MA, or Partisil-50DS, Whatman Inc., Clifton, NJ). The support phase was silica and the composition of the 23 mobil phase changed linearly over a 20 minute analysis period from 100 percent methanol/water (70/30) to 40 percent. methanol/water and 60 percent NmM KHZPOA buffer. Column flow was usually 1.5ml/min and the column was run at ambient temperature. Absorbance of the column eluate was continuously monitored at 259nm (Waters Model uuo Absorbance Detector) and recorded with a Waters data module. This gradient and flow typically eluted adenosine at a retention time of 18.” minutes. The adenosine peak in an unknown sample was identified by correspondence of its retention time with that of a known adenosine standard and the absence of an adenosine peak in samples treated with adenosine deaminase to remove adenosine. Sample peak heights were directly measured in mm and compared to the corresponding adenosine standard peak height. In each experiment a sample "spiked" with adenosine (approximately 0.3nM) and three samples with adenosine deaminase added were collected simultaneously with other blood samples as described in the previous section. These samples were then used to assess the reliability of the assay. The adenosine peak of the "spiked" sample was compared to that of the ”unspiked" sample. If the increase in peak height of the adenosine "spiked" sample was less than 80 percent of what it should be, then all the values for that experiment were rejected. Likewise, if the peak heights of any adenosine deaminase samples were greater than 30 percent of the peak height of their corresponding blood samples, all the sample values for that particular experimental alteration were rejected. When an adenosine deaminase sample exhibited a measureable peak less than 30 percent of its paired blood sample, the residual peak was 24 subtracted from the peak heights of all the samples in that particular experimental alteration. C. Prostacyclin 1. Sample Collection Blood was collected to determine whether prostacyclin was involved in the vasodilation associated with hypoxia and/or hypercapnia. Approximately 2ml were collected into precooled 9ml vacutainer tubes containing the potassium salt of the chelating agent ethylenediamine tetraacetic acid, EDTA, and centrifuged for five minutes at 2800rpm (1360xg) and ”CC. One ml of plasma from each sample was pipetted into a precooled polyprOpylene tube (12mm by 75mm). Samples were stored at -2o°c until assayed. 2. Radioimmunoassay (RIA) A detailed description and the validation of the prostacyclin RIA are given in the Appendix. Briefly, the stable breakdown product of alpha (6-keto-PGF prostacyclin, 6-keto-prostaglandin F alpha). was 1 1 analyzed by radioimmunoassay. All reagents were diluted in a 0.1M phosphate buffer. Rabbit antiserum, 100ul (Seragen, Boston, MA). specific for 6-keto-PGF1 alpha was added to 100ul of unknown sample. Tritiated 6-keto-PGF1 alpha, 100ul (specific activity 120.0 Ci/mmol, New England Nuclear, Boston, MA), was also added and the tubes were mixed, incubated for one hour at 214°C and then incubated at 11°C for 18-211 hours. Tubes with known amounts of 6-keto-PGF1 alpha (0.00A8-2.5ng/100ul, U51787. The Upjohn Co., Kalamazoo, MI) were included in each assay. These latter tubes served as the standard curve and were used to calculate the quantity of prostaglandin in the unknown 25 samples. All standards and samples were run in triplicate. After incubation, the tubes were placed on ice and lml of a precooled charcoal:dextran (0.51:0.051) suspension was added to each tube for 12 minutes“ ‘Then the tubes were centrifuged at 3000rpm (2000xg) in as refrigerated (NOC) centrifuge (Beckman model J-6B, Palo Alto, CA) for 12 minutes. One ml of the supernatant was pipetted, added to 15ml of scintillation fluid (Packard, Downers Grove, IL), and counted for ten minutes or a statistical accuracy of less than 2.0 percent on a Tri-Carb 3000 liquid scintillation counter (Downers Grove, IL). Quality control samples were run in each assay. Within- and between— assay variations were 7.5 percent and 11.9 percent, respectively, for a 100ul plasma volume. Serial dilutions of dog plasma were parallel to the standard curve. Dog plasma stripped of endogenous prostacyclin with a charcoal:dextran suspension (5.01:0.5i) had no measurable 6-keto-PGF1 alpha. Known amounts of 6-keto-PGF alpha added 1 to "charcoal stripped" plasma were used to determine the accuracy of the assay. 3. Infusions 2f Prostacyclin (P012) To determine the vasodilator effect of prostacyclin, it was administered exogenously. In four animals, prostacyclin (U53217A, The Upjohn Co., Kalamazoo, MI) was infused after the hypoxic and hypercapnic eXperiments. A 200ng/ml solution of prostacyclin was delivered by an infusion pump (Harvard Apparatus Co., Millis, MA) into the brachial artery at three successive infusion rates (0.999, 1.23, and 2.u7ml/min). Steady states, with respect to perfusion and blood pressures, were usually attained within four minutes from onset of each prostacyclin infusion rate. Blood samples were then drawn from all four sampling 26 sites to determine the concentration of prostacyclin entering and draining the forelimb. Resistances were determined and correlated with prostacyclin concentrations. D. Statistical Analysis The data were statistically analyzed by a two-way' analysis. of variance and the paired Student's t test. Significance was taken at p<0.05 or p<0.01 (26). III. RESULTS A. Perfusion Pressure, Resistance, and Blood Flow Vascular changes that occurred in the isolated forelimb in response to alterations in gas tensions are shown in Figures 2 and 3. Mild hypoxia, severe hypoxia and hypercapnia all significantly (p<0.01 or p<0.05) decreased the perfusion pressure and resistance in the vascular bed of the forelimb (Figure 2). Severe hypoxia produced the most dramatic changes: perfusion pressure decreased from a mean of 113mmHg before hypoxia to a mean of 79mmHg after severe hypoxia because resistance decreased from a mean of 1.05mmHg/ml/min to a mean of 0.72mmHg/ml/min. These changes indicate vasodilation in the whole forelimb because there was no redistribution of blood flow between the muscle and skin. This is indicated by consistent blood flows in the brachial and cephalic veins before and after severe hypoxia (Figure 3). B. Adenosine The adenosine concentrations measured in the forelimb plasma during alterations in gas tensions are shown in Figure A. Mild hypoxia, severe hypoxia, and hypercapnia did not change the adenosine production across the vascular bed of the forelimb. Adenosine concentrations were approximately 0.1uMolar in all of the vessels and did not increase from artery to vein with hypoxia or hypercapnia. 27 28 :mcsm w. wmsmcuwo: nsmuucso Awswwo: csmuucsm Sauna «so m umwwma mncom:«.u a «man too cmswosSma no ooavmsm «so oosnsoH powwoa tan: «am mxcmswaoswmw mwwmsmwwon. auvAo.om. souvAo.od. z aJnBId 29 Perfusion Pressure (mmHg) 0 O O I I I I I .s R) O I x x \ HWO HS 0 8L L’ 8L / * + * * UdH O Resistance (mmHg/mI/min) .° .° 1" O uh on N I I I I I I O; j—A §:‘ \\7 1* I *- HS 3 8L / sat UdH O 30 Figure _3_ Blood flow through the brachial, BV, and cephalic, CV, veins during control (C,15-20%02:5%C02:75-801N2) and severe hypoxia (SH,O%02:5SC02:95%N2). Data represent means : SEM, n=6. A paired Student's t test was performed to compare control and severe hypoxia in each vessel. r yZ/eww I. n . a. Ease: 25E 32 mwmcso_w nozoosnsonwosu ow moozouwso »: vwmuam tmcos Aczopmsv acswsm oosnwow An.HmImouomumanomuqmlmouzmv. appa swuoxwm Az:.wuomnmuoomuoouzmv. uo u asmOSHmp manos< w< u osmovwmp vowsoa macaosn.u a noun :mu cosmosaoa so ooacmso «so maosouwso oosoosnsmnwo: M: «so oosasow 0.05) than the concentrations in the vessel supplying the forelimb (i.e. brachial artery). The increase in plasma 6-keto-PGF1 alpha concentrations from the femoral artery to the forelimb vessels represent a constant production. of prostacyclin by' the lungs (21). Average concentrations in the forelimb veins ranged from 1.69 to 2.18ng/ml plasma. Prostacyclin was infused into the brachial artery of the forelimb and plasma samples were drawn to determine the concentrations of 6-keto-PGF1 alpha at the four sampling sites (Figure 6). The concentrations measured in the systemic circulation, as well as in the veins draining the forelimb, were significantly different from the level of 6-keto--P(3F1 alpha of the brachial artery in all groups (p<0.05 or p<0.01). With infusion, the mean values in the veins rose from a control concentration of 1.88ng/ml plasma to 3.32, 5.1”, 8.81ng/ml plasma with increasing infusion rates of prostacyclin (98.8, 162.8 and 494ng/m1, respectively). The changes in systemic blood pressure and forelimb perfusion pressure and resistance during prostacyclin infusion 35 gm. oosoososmnwosu om mixo¢01usounmmwmsaws ms oposm Asmxap upmuamv acswsm oosnsow Ao.pmlmouomumanomuqmlmoazmv. apps szo<> 2mm cosmossoo no ooaomso «so olxonoLuom.d mwosm oosoosasmawosu mosouu vowsoa m«caos«.u « «ou« zmu cosmosaoa «o ooacmso «so oos«sow msa omos ow «so Mswcuwo: msoccu. eu vo.om. emu qu.os. 38 ._0I m l @353: 0* t~0m~mo m> m< o< I» m> m< o< n> m> m< o< n> m> m< o< 002.30.. .353: d .3253: n .3253: a Aomb 393.3 2 mn.msm\3.3 wflmcno o . 3653933. 39 are shown in Figure 7. The mean systemic blood pressure decreased from 83 to 62mmHg with increased concentrations of prostacyclin. Although not statistically significant, due to the small sample size, this hypotensive effect has been documented in the literature (1). Similarly, perfusion pressure and resistance in the forelimb declined with increased prostacyclin infusion rates. Perfusion pressure decreased from 135mmHg perfusion pressure to lOlmmHg and resistance declined from 1.04 to 0.662mmHg/m1/min. 40 mwmcso.H m mizow >zo<> sou cuoa «o coaumso s«wsoa< wooop