ABSTRACT STUDY OF THE AVIAN CEREBROSPINAL FLUID USING BRAIN VENTRICULAR PERFUSION By Douglas Kroulik Anderson Brains of anesthetized chickens were perfused from the left lateral ventricle to the cisterna magna with an artificial chicken cerebrospinal fluid (CSF) containing radio-iodinated human serum albumin (RIHSA) or inulin, 22Na, 42K, 45Ca, 14C-glucose, 14C-creatinine and unlabelled crea- tinine. Inflow (Vi) and outflow (90) rates and concentrations of all test molecules were measured. The steady-state clear- 22 ances of RIHSA (C Na (CNa) and creatinine (Ccr) from RIHSA)’ the perfusate were calculated at various intraventricular pressures. CRIHSA increased and outflow-inflow (VG-Vi) decreased linearly (with equal but opposite slopes) with increasing intraventricular pressures, suggesting CRIHSA was a measure of CSF bulk absorption (Va). CSF formation rate (Vf) calculated as the algebraic sum of CRIHSA and VO-Vi, was approximately 1.4 ul/min., and was independent of intra- ventricular pressure. Resistance to 9a (4545 cm- min/ml) was 12-350 times that reported for mammals and 2.5 times Douglas Kroulik Anderson that reported for turtles. This high resistance may indi- cate either a lack of valve-like channels in the arachnoid villi (as described for mammals) or high resistance pathways in the arachnoid membrane. Total clearance of sodium and creatinine was divided into a pressure-dependent component related to Ta and a pressure-independent component termed an efflux coefficient (KD). The steady-state efflux coefficients for creatinine (KDcr)’ 42 45 K (KDK), and Ca (KDCa) were not affected by their perfusion inflow concentrations suggesting that the non—bulk removal of all three molecules from chicken CSF is by simple diffusion paralleling,previous studies in various mammalian species. The large KD for creatinine (relative to KDNa’ KDCa’ and KDK) suggests active transport may be involved in creatinine efflux. Ventricular CSF volume was estimated from the distri- 22 14 bution volume of RIHSA, Na and C-glucose (determined by integrating the outflow concentration of these molecules with respect to time) and total CSF volume was estimated from the maximum volume of CSF that could be withdrawn from the cisterna magna. Ventricular volume was 140 pl; total CSF volume was 350 pl. Brain spaces for RIHSA, 22Na, 45 42 Ca and K were estimated from brain residual radioactivity after 1-6 hours of perfusion. The smallest brain space was 4 percent of brain weight for RIHSA and the largest was Douglas Kroulik Anderson 91 percent for 42 K. Magnitude of brain space was dependent on several factors: molecular size; magnitude of brain up- take; and extent of uptake by brain cells. Perfusion of 0-40 mEq/L of potassium through the chicken cerebral ventricles was without effect on systolic, diastolic or mean blood pressure. Ventricular perfusion with 0-5 mEq/L calcium produced similar results for diastolic and mean blood pressure but systolic blood pressure was signifi- cantly depressed at elevated calcium concentrations. With the exception of the systolic response to calcium, these results do not confirm previous studies in mammals. It is suggested that the lack of response in chickens is due to anesthetic interference with the sensitivity of the vasomotor center to ionic changes in CSF. STUDY OF THE AVIAN CEREBROSPINAL FLUID USING BRAIN VENTRICULAR PERFUSION By Douglas Kroulik Anderson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Physiology 1972 ACKNOWLEDGMENTS I would like to express my sincere gratitude to Drs. R. K. Ringer and S. R. Heisey, my major advisors, for their friendship, encouragement, guidance, and aid in the preparation of this dissertation. I would also like to thank the remaining members of my committee, Drs. W. D. Collings, J. R. Hoffert, and E. P. Reineke for their advice and assistance with this research effort. I would like to express my appreciation to Miss Nancy Turner for her patience in the typing of the rough drafts of this thesis and to Ronald Rook for his assistance with parts of this study. Special thanks are extended to my wife, Elizabeth, and children, Michael and Stacey for their love, patience and understanding throughout my doctoral program. ii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . . . . . . . . . . . vi INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 1 LITERATURE REVIEW . . . . . . . . . . . . . . . . . . S Cerebrospinal Fluid Formation and Absorption . . . 5 Molecular Exchange Between Blood, Brain, and Cerebrospinal Fluid . . . . . . . 15 Cerebrospinal and Brain Extracellular Fluid Volumes . . . . . . . . 26 Effect of Ionic Changes in Cerebrospinal Fluid on Cardiovascular and Respiratory Functions . . . 35 STATEMENT OF THE PROBLEM . . . . . . . . . . . . . . . 41 METHODS AND MATERIALS . . . . . . . . . . . . . . . . 42 A. Animals . . . . . . . . . . . . . . . . . . . . 42 B. Anesthesia . . . . . . . . . . . . . . . . . . 42 C. Surgical Procedures . . . . . . . 44 D. Cerebral Ventricular Perfusion Technique . . . 49 E. Calculations . . . . . . . . . . . . . . . . 55 F. Cardiovascular Measurements . . . . . . . . . . 62 G. Statistical Analysis . . . . . . . . . . . . . 63 RESULTS . . . . . . . . . . . . . . . . . . . . . . . 64 Cerebrospinal Fluid Formation and Absorption . . . 64 Molecular Flux from Cerebrospinal Fluid to Brain and Blood . . . . . . . . . . . . . . . . 67 Mass Balance Determination . . . . . . 88 Cerebrospinal Fluid Volumes and Brain Spaces . . . 91 Effects of Altered Cerebrospinal Fluid Potassium and Calcium Concentrations on Cardiovascular Functions . . . . . . . . . . . . . . . . . . . . 95 iii DISCUSSION . . . . . . . . . Cerebrospinal Fluid Formation and Absorption Molecular Flux from Cerebrospinal Fluid to Brain and Blood . . . . . . . Cerebrospinal Fluid Volumes and Brain Spaces Effects of Altered Cerebrospinal Fluid Potassium and Calcium Concentrations on Cardiovascular Functions . . . . SUMMARY APPENDICES I. PREPARATION OF ARTIFICIAL CEREBROSPINAL FLUID . . . . . . . . . . . . . . II. LIQUID SCINTILLATION COUNTING III. FLAME PHOTOMETRY IV. ATOMIC ABSORPTION SPECTROSCOPY V. CREATININE ASSAY VI. INULIN ASSAY . . . . . . . . . . . . . . VII. SAMPLE CALCULATION OF CEREBROSPINAL FLUID FORMATION AND ABSORPTION RATES, OUTFLUX COEFFICIENT, AND OUTFLUX RATE . . VIII. EXTENDED FORTRAN PROGRAM FOR THE CALCULATION OF MASS BALANCE . . . . . . . . BIBLIOGRAPHY . iv Page 102 102 105 110 113 119 122 125 128 131 134 137 140 143 147 Table LIST OF TABLES Page CerebrOSpinal fluid (CSF) formation and turnover rates in different vertebrate species . . . . . ll Cerebrospinal fluid volumes in different vertebrate species . . . . . . . . . . . . . . . 28 Effect of perfusion time on the efflux coefficients for sodium, calcium, potassium, and creatinine . . . . . . . . . . . . . . . . . 87 Ventricular perfusion mass balance for RIHSA, 2zNa, 45Ca, and 42K . . . . . . . . . . . . . . 89 Distribution of RIHSA, 22Na, 45Ca, and 42K in the chicken brain . . . . . . . . . . . . . . . 94 Effect of anesthesia and cerebral ventricular perfusion on blood pressure and heart rate in ChiCkenS O o a o o o o o o o o o o o o o o o 96 Figure 10. 11. LIST OF FIGURES Diagram of brain ventricular perfusion system in reference to a saggital midline section of a chicken head . . . . . Results of a single perfusion experiment from the left cerebral ventricle to the cisterna magna in a chicken . . . . Mean cerebrospinal fluid (CSF) formation rate, RIHSA clearance, and difference between ventricular outflow and inflow rates as a function of intraventricular pressure . 22Na non-bulk clearance, total cular pressure . . . . . . Creatinine non-bulk clearance, total crea- tinine clearance, and creatinine bulk 22 Na clearance, and 22Na bulk absorption versus intraventri- absorption versus intraventricular pressure . Creatinine outflux coefficient plotted as a function of perfusion inflow creatinine concentration . . . . . . . Creatinine outflux rate plotted as a function of perfusion inflow creatinine concentration Potassium outflux coefficient versus perfusion inflow potassium concentration Potassium outflux rate as a function of perfusion inflow potassium concentration Calcium outflux coefficient plotted as a function of perfusion inflow calcium concentration . . . . . . . Calcium outflux rate as a function of perfusion inflow calcium concentration vi Page 47 52 66 70 72 75 77 80 82 84 86 Figure Page 12. Mepn distribution volumes of RIHSA, 22Na and 4C-glucose in cerebral ventricular system of the chicken . . . . . . . . . . . . . . . 93 13. Experimental systolic, diastolic, and mean blood pressures (expressed as percent of control blood pressure) plotted as a func- tion of perfusion inflow potassium concentration . . . . . . . . . . . . . . . . 98 14. Experimental systolic, diastolic, and mean blood pressures (expressed as percent of control blood pressure) versus perfusion inflow calcium concentration . . . . . . . . 101 vii INTRODUCTION The central nervous system (CNS) during development assumes a tubular configuration and retains the character- istic of being a hollow organ throughout life. This neutral tube evolves into the various brain ventricles which contain cerebrospinal fluid (CSF), a low protein secretion derived from blood plasma (Millen and Woollam, 1962; Davson, 1967). CSF is also found in the subarachnoid spaces which cover the entire surface of the CNS. The circulation of CSF has been well described for mammals (Davson, 1967; Millen and Woollam, 1962; Truex and Carpenter, 1969). CSF flows from the two lateral cerebral ventricles through the foramina of Monro into the third ventricle. From the third ventricle the CSF passes through the Aqueduct of Sylvius into the fourth ventricle. CSF leaves the fourth ventricle by way of 3 foramina, a medial foramen of Magendie and two lateral foramina of Luschka and enters the subarachnoid spaces surrounding the brain and spinal cord. Total CSF volume in humans has been estimated to be 140 m1 (Millen and Woollam, 1962). Since the studies of Frazier and Peet (1914) and Dandy (1919) on dogs, it has been generally accepted that CSF is a secretion from specialized vascular structures (choroid plexuses) which, along with a lining of ventricular ependyma, project into all four brain ventricles. This has been confirmed by de Rougemont and his co-workers (1960) who collected newly formed CSF directly from the choroid plexus of cats. Recent evidence indicates that in mammals, the ventricular ependyma (Bering and Sato, 1963; Milhorat, 1969; Pollay and Curl, 1967; Welsh, 1963) and the cranial subarachnoid spaces (Bering and Sato, 1963; Sweet and Locksley, 1953; Wallace and Brodie, 1940) may also play a role in the elaboration of CSF. The studies of Frazier and Peet (1914) and Dandy (1919) indicate that, in dogs, CSF is reabsorbed into venous blood mainly (if not solely) from the subarachnoid space. Weed (1914) pr0posed that the arachnoid villi (evaginations of the arachnoid membrane into the dural venous sinuses) are the structures responsible for draining CSF into blood. The importance of the arachnoid villi in CSF drainage was recently confirmed by Welsh and Friedman (1960) who described these structures in monkeys as a labyrinth of small tubes which establish an open connection between CSF and venous blood. They showed that these villi function as valves which Open when CSF pressure is higher than sagittal sinus pressure and close when the dural sinus blood pressure exceeds CSF pressure. One function of CSF is mechanical, i.e., it serves as £1 fluid cushion for the CNS (Millen and Woollam, 1962; Truex and Carpenter, 1964). In addition CSF may: (1) serve as a nutrient source for the brain and spinal cord and aid in waste removal (Millen and Woollam, 1962; Truex and Carpenter, 1964); (2) carry the hypophyseal hormones (Millen and Woollam, 1962); and (3) play a role in the control of respiration (Pappenheimer g£_gl., 1965). As for most non-mammalian species, inadequate data are available to describe CSF formation and reabsorption characteristics in birds. From a comparative physiological standpoint, and as an extension of a recent study on the normal composition of chicken CSF (Anderson and Hazelwood, 1969) this information on avian species is important. Further, the effect of changes in CSF concentrations of K+ and Ca++ on blood pressure and heart rate has been studied in several mammalian species (reviewed by Tschirgi, 1960 and Winterstein, 1961), however no studies of this type have been performed with birds. The control of the avian cardiovascular system differs from that of mammals. Although birds have a carotid sinus homologue, they do not exhibit a carotid sinus reflex (Heymans and Neil, 1958; McGinnis and Ringer, 1966). Sturkie (1965) indicates that baroreceptor mechanisms are present in birds; however, their anatomical locations are unknown. Since birds differ from mammals in the control of their blood pressure and heart rate, it is important to determine if birds and mammals vary in their response to changes in the ionic composition of CSF. Although direct neutral con- trol of blood pressure in chickens has been postulated (Rodbard and Tolpin, 1947), no evidence is available to determine the presence or to establish the location of a vasomotor center (Sturkie, 1965). In general, perfusing or injecting solutions with a high K+-Ca++ ratio (due either to high levels of K+ or low levels of Ca++) into the CSF system of mammals results in an elevated arterial pressure concomitant: with a slowing of the heart rate (Tschirgi, 1960). These cardiovascular responses are attributed to the direct action of elevated K+ or reduced Ca++ levels on the medullary cardiovascular centers (Tschirgi, 1960; Winterstein, 1961). Changes in blood pressure and heart rate in chickens during ventriculocisternal perfusion with CSF having an altered K+-Ca++ ratio might be taken as pre- sumptive evidence for the location of vasomotor and cardiac centers in the avian brain in proximity to the path of the CSF perfusate. LITERATURE REVIEW The purpose of this review is to present studies in several species (mostly mammalian) on: (1) cerebrospinal fluid (CSF) formation and reabsorption rates; (2) molecular exchange between CSF, brain and blood; (3) determination of the brain ventricular and extracellular fluid volumes; and (4) the effect of changes in CSF K+ and Ca++ concentrations on blood pressure, heart rate, electrocardiogram (EKG), electroencephalogram (EEG) and respiration. Cerebrospinal Fluid Formation and Absorption Frazier and Peet (1914) were the first workers to quantitatively determine the rate of CSF production. They sutured blunt cannulas (connected to a rubber tube and graduated pipette) into the cisterna magna of anesthetized dogs. The dog was placed on an inclined plane, head down, with the level of the outflow cannula below the head so that all the fluid entering the cisterna magna would be collected. The rate of formation in dogs was found to vary widely under normal conditions, the average being 0.231 ml/min. Frazier and Peet also obtained an estimate of the rate of CSF reab- sorption by determining the rate of appearance of phenol- sulphonephthalein (PSP) in the urinary bladder after a known quantity of this dye had been injected into the ventricles. They assumed CSF is reabsorbed at a rate comparable to PSP. clearance by the kidney indicating that 0.4-0.5 percentmo ma.o om.o ON pmu Aomfi .Hazo w amfifloa om.o ma panama mama .hpomo mm.o N.N “my Hamfl .Hmmguflz w Ammumz mH.o oo.o a.H mfiphsp .. .. .. v thnm :oEoH : Iltll : m xhwzm ownsa mama ..Hm pm paommo mo.o v :mfimmow zqwmm mononomom opmh msxoflm :«E\H: mofloomm uo>ocpse mmu machono ma non :fiE\a: :ofipmsaom mmu .mowoomm ouwnnouho> pcohommflw a“ moumu ho>o:HSu paw :ofiumehow pwsHm HmcflmmOAEouoo .H ofiamg 12 acetazolamide. He indicated that CSF secretion from the cho- roid plexus is almost completely stOpped under the influence of intravenously administered or topically applied acetazola- mide. Welch cites other studies which show that total CSF production (as measured by inulin dilution) is decreased by only about 50-60 percent under the influence of this drug. Since acetazolamide essentially stopped choroidal production of CSF in_vi£rg while only cutting CSF production in half in 2312) he concluded there was a dual source of CSF forma— tion, from the-choroid plexuses and from the surrounding nervous tissue. Cerebral ventricular perfusion can also be used to demonstrate extra-choroidal CSF formation. Pollay and Curl (1967) using an aqueductal-anterior fourth ventricle perfusion system in the rabbit calculated a mean rate of CSF formation (determined by inulin dilution) originating from the ventricular ependyma of 0.33 ul/min per cm2 of ependymal surface. The total area of ventricular ependyma in the rabbit was 12.8 cmz giving a CSF formation rate attributed to the ventricular ependyma of 4.23 ul/min which is approximately 33 percent of the total intraventricular CSF formation rate in the rabbit. Milhorat (1969) found that CSF production was decreased by only 26 percent in bilaterally plexectomized rhesus monkeys (studied by ventriculo—aqueductal perfusion) as compared to normal nonplexectomized controls. He con- cluded that in the monkey the choroid plexus is not the sole or even major source of CSF, this being either a secretion 13 of the ventricular ependyma or a product of cerebral metabo- lism which enters the ventricular system across the ependymal lining. Bering and Sato (1963) using ventriculocisternal and subarachnoid:cisternal perfusion, found in the anesthe- tized dog that about 43 percent (0.020 ml/min) of total CSF production was by the cranial subarachnoid space. However, Hammerstad §£_al. (1969) using a cisternal-lumbar perfusion in anesthetized cats could not detect any CSF formation from the spinal subarachnoid space. The ventriculocisternal perfusion technique does not only allow the determination of the sources and rates of CSF secretion but also possesses capabilities for studying the relationship between the formation of CSF and intraventricular pressure. As mentioned previously, (Heisey gt_al. (1962) dem- onstrated in the goat the independence of CSF formation and acute changes in intraventricular pressure. This relation- ship between CSF formation and intraventricular pressure has been well established in a variety of species, notably the turtle (Heisey and Michael, 1971), cat (Katzman and Hussey, 1970), human (Cutler, et_al., 1968), and dog (Bering and Sato, 1963). Bering and Sato further state that the rate of CSF formation is unaffected by the development of hydrocephalus. Recently Sahar (1970) demonstrated, in individual anesthe- tized cats with kaolin-induced obstructive hydrocephalus, that there was a decrease in the CSF formation rate of 3.5 percent/cm H20 as perfusion pressure was elevated. 14 Their stated purpose for looking at individual animals was that in studies previous to theirs, conclusions concerning the influence of perfusion pressure on CSF formation rate have been based only on statistical evaluation of whole groups. Thus, studying individual animals would allow paired comparisons. They propose that the reduction in CSF produc- tion was due to an adverse effect of increased intraventricu- lar pressure on the structure of the chorid plexuses and ventricular ependyma. In a later study Sahar g£_al. (1971) found a 30 percent.reduction in CSF production in individual anesthetized kaolin-induced hydrocephalic dogs compared with normal anesthetized dogs. They found that this reduction in CSF production is not as great as that seen in their previous study on cats because the choroid plexuses in the hydrocephalic dogs were not damaged. It appears that although the increased intraventricular pressure had a deleterious effect on the structure of the ventricular epen- dyma in dogs, it did not reduce CSF production by the choroid plexuses of these animals. The technique of cerebral ventricular perfusion has been instrumental in the greater quantitation of studies on CSF formation and absorption rates in a variety of animal species. The versatility of this technique also allows the study of CSF secretion at various intraventricular pressures in either acute or chronic pathological (hydrocephalus) situations. Measurements of the ventricular and brain 15 extracellular fluid volumes, studies of the movement of various substances into and out of the CSF system, and inves- tigations of peripheral autonomic function by ventriculo- cisternal perfusion (all to be discussed later) further emphasize the adaptability of this technique. Molecular Exchange Between Blood, Brain and Cere rospinal Fluid Determination of the CSF formation and absorption rates is integral to the study of molecular flux between CSF, brain and blood. With knowledge of these parameters appro- priate permeability coefficients can be devised that take into account material influx into the ventricular system with freshly formed CSF or subtract the material outflux from the ventricles with CSF absorbed in bulk. These per- meability coefficients have been used to compare the relative rates of movement of a variety of materials across the ven- tricular wall and have resulted in greater quantitation in the study of the blood-brain and blood-CSF-barriers. Movement from blood to CSF and brain Early studies with intravenously injected dyes showed that the brain was spared from staining while most of the other body tissues were stained. It was this restricted passage of dissolved dyes out of the blood into brain tissue and CSF that gave rise to the concepts of 16 blood-brain and blood-CSF barriers. However, if dyes were injected directly into CSF, the whole brain became heavily stained (Tschirgi, 1960; Davson, 1967). Thus, while there appears to be a barrier between blood and nervous tissue and CSF, this barrier can be circumvented by direct injection into CSF as there does not appear to be a barrier between CSF and brain. Davson (1967) has investigated the penetration from blood of various non-electrolytes into the brain and CSF of rabbits. The rate of penetration decreased in the following order: ethyl alcohol > prOpyl thiourea > ethyl thiourea > methyl thiourea > thiourea > creatinine. Davson states that the decreases in the rates of penetration are related to the decreasing lipid-solubility of the substances. One way of explaining the permeability character- istics exhibited by the blood-CSF and blood-brain barriers to certain lipid-insoluble molecules is by carrier mediated transport which includes both active transport and facili- tated transfer. When a solute diffuses down an electro- chemical gradient, the process is described as passive diffusion. When the material moves against an electrochemi- cal gradient, exhibits saturation kinetics, requires metabolic energy and can be competitively inhibited, the process is termed active transport. If the transport system exhibits saturation kinetics and competitive inhibition but the transport is not against an electrochemical gradient and 17 is not energy dependent, the process is termed facilitative transfer. Fishman (1964) reported on the movement of sugars between blood and CSF in anesthetized dogs. He found that CSF glucose concentrations approached a maximum despite increasing intravenous glucose loads, suggesting saturation kinetics. Competitive inhibition between glucose and 2- deoxyglucose was also demonstrated. Fishman concluded there was a carrier transport system for glucose in the membranes separating blood and CSF. Bradbury and Davson (1964) also found evidence for a saturable carrier transport system for D-glucose and D-xylose into and out of the CSF of anesthe- tized rabbits. These workers found no evidence for transport against a concentration gradient or energy requirement sug- gesting a facilitative rather than an active transport system for these monosaccharides. One of the first quantitative studies of the move- ment of ions across the blood-brain and blood-CSF barriers was by Wallace and Brodie (1940). These workers administered bromide, iodide and thiocyanate by iv injection in anesthe- tized dogs. They found that the uptake of these ions by the CNS and CSF was slow and restricted in comparison to other tissues and concluded they must cross a barrier "which offers some selective hindrance to their course" in passing from plasma to the brain extracellular fluid (ECF) and from there into CSF. Greenburg g£_al. (1943) studied the permeability of the blood-CSF barrier to a number of radioactively 18 labelled ions in anesthetized dogs. They found that the rate of increase in concentration of the labelled ion in CSF, following intravascular administration was in the order potassium > sodium > bromide > rubidium > strontium > phosphate > iodide. These workers indicated that the rate of accumulation in CSF of the injected ions (both positively and negatively charged) is selective and is a slow process in comparison to other tissues. They concluded that CSF formation must be a secretory process rather than passive diffusion. Katzman and Leiderman (1953) found that the rate 42K between the plasma and brain of rats of equilibration of was slow. In addition they found that despite variations in the plasma potassium levels, the potassium flux into brain from plasma is approximately the same, suggesting that potassium movement into brain is carrier mediated; a finding they confirmed in cats several years later (Katzman g£_al., 1965). Graziani gt_al. (1967) studied the calcium flux from blood to brain and CSF (as measured by 45Ca flux) in anes- thetized cats during ventriculocisternal perfusion. These workers found that when the serum calcium levels were varied, a component of the calculated permeability coefficient for 45Ca from blood to CSF and brain was reciprocally related to serum calcium concentrations indicating an active or carrier- mediated process. Another smaller component of the coeffi- cient was constant which is consistent with passive diffusion. ‘When ouabain was added to the perfusate both CSF formation 19 rate and calcium influx declined suggesting a calcium influx component related to CSF formation. However, when aceta- zolamide was added to the perfusate, CSF formation declined but a component of calcium influx continued independent of the reduced CSF formation suggesting that calcium can enter CSF by ways other than with freshly formed fluid. Bito (1969) presented evidence that the dog cerebral cortex ECF has a low potassium and high magnesium concentration even when compared with the concentration of these cations in cisternal CSF. He concludes that these brain ECF concentra- tions of potassium and magnesium can not be maintained by the secretory activity of the choroid plexuses and a passive diffusional barrier between blood and brain but necessitate the existence of an active transport function across the blood-brain barrier. Any interpretation of ionic distributions and fluxes between CSF, brain and blood requires knowledge of any elec- trical potential differences existing between these com— partments. Held g£_al. (1964) determined the steady-state electrical potential difference between cisternal fluid and jugular venous blood averaged +6.5 mV in unanesthetized goats and varied from -2 to +7 mV in anesthetized dogs. They found there was an inverse linear relationship between the pH of arterial blood and the CSF potential changing from +15 mV at pH 7.1 to -3 mV at pH 7.6. They also observed an inverse relationship (slope = 1 mV per mEq/L) at constant 20 arterial pH between the potassium concentration of fluid perfusing the cerebral ventricles and the CSF potential. In addition, they found that the rate of CSF secretion was unaf- fected during large changes in CSF potential. These workers concluded that the CSF potential must contribute to the exchanges of all charged particles between CSF, brain and blood and must play a role in determining the steady-state ionic composition of CSF. They further suggest that the CSF potential is derived from ion transport across the ven- tricular ependyma and is not associated with CSF formation. Movement from CSF to blood and brain The removal of substances from CSF depends upon several mechanisms: bulk flow of fluid via the arachnoid villi into blood, active transport of substances into nervous tissue or blood, and diffusion of substances into nervous tissue or blood. As discussed previously, bulk absorption refers to the pressure-dependent removal of CSF with its total contents through the one-way valve system of the arachnoid villi. This is mainly how large, non-diffusible nmlecules (e.g., proteins, inulin, dextran) exit the CSF system. Since there is no filtration across the arachnoid villi, one component of the total removal of all substances from the CSF compartment will be with the bulk absorption of fluid through the arachnoid villi into blood. 21 A number of substances have been shown to be actively transported from the CSF system. Pappenheimer g£_§l. (1961) presented evidence for the active transport of Diodrast and phenolsulphonephthalein (PSP) from CSF to blood in anesthe- tized goats perfused from one lateral ventricle to the cisterna magna. These workers found that Diodrast moved from CSF to blood against a concentration gradient, was com- petitively inhibited by para-aminohippuric acid (PAH) and PSP and exhibited saturation kinetics. They presented evi- dence implicating the choroid plexus of the fourth ventricle as the site of the active transport. Snodgrass g£_al, (1969) perfused the ventriculocisternal system in anesthetized cats to study the transport of neutral amino acids out of the CSF system. They found that the exit of L-leucine and cyclo- leucine from CSF exhibited saturation kinetics. In contrast, the movement of a-aminoisobutyric acid (AIB) and L-alanine showed no saturation kinetics. These workers concluded that neutral amino acids were not cleared from CSF by bulk absorp- tion or diffusion alone, but for some amino acids there was a saturable transport out of the CSF. Murry and Cutler (1970) studied the rate of clearance of glycine from the CSF of anesthetized cats during ventriculocisternal and ventri- culolumbar perfusion. They found that glycine was cleared from the ventricular system and spinal subarachnoid spaces by a saturable mechanism, suggesting carrier-mediated trans- port, and also demonstrated that the choroid plexuses and 22 periventricular thalamic tissue could concentrate glycine to about 10 times that of the surrounding medium. Several workers have found that various anions are transported out of the CSF. Pollay (1966) has indicated that active transport of thiocyanate from CSF is responsible for maintaining the observed concentration gradient of thiocyanate between brain and CSF. Destroying this transport process results in a more accurate estimate of the brain thiocyanate space (see page 32 of this review). Cutler g£_al. (1968), using ventriculocisternal perfusion in anesthetized cats, showed that iodide and sulfate were cleared from CSF against a concentration gradient, exhibited saturation kinetics and could be competitively inhibited. In addition, they found that sulfate was cleared from the CSF of adult cats at a rate 3 times greater than that from kittens and concluded that transport from CSF was more efficient in adult cats than in kittens. Hammerstad (1969) used a technique of cisternal-lumbar perfusion in anesthetized cats to study the transport of iodide and sulfate out of the spinal subarach- noid space. They found that iodide was cleared from the spinal subarachnoid spaces by a saturable process, against a concentration gradient, which could be competitively inhibited by thiocyanate. In contrast, sulfate removal from CSF was unaltered by changing the perfusate sulfate concen- tration. These workers concluded that iodide was transported 23 out of the spinal subarachnoid fluid by a carrier-mediated process whereas sulfate exited by passive diffusion. Many substances other than sulfate have been shown to leave the CSF by a process of simple or passive diffusion. Davson g£_al. (1962) found a "relatively high rate of escape" from the CSF of lipid-soluble substances such as ethyl- thiourea as compared with 24Na. The rate of inulin loss was found to be small and these workers indicated that inulin loss was probably almost exclusively by way of the arachnoid villi. Pappenheimer g£_al. (1961) reported that after inhibi- tion of active Diodrast transport a passive component of transfer of this molecule was revealed. Studies with cre— atinine, fructose and inulin showed that passive movement of these molecules from CSF is by both bulk absorption and diffusion at rates comparable with diffusion rates from the capillaries of 1 gm of skeletal muscle. Heisey g£_al. (1962) introduced increased quantitation in the study of molecular movement out of the CSF by deriving a diffusional permeabil- ity coefficient, KD, that accounts for movement of substances out of the CSF by means other than by bulk absorption. These workers compared the steady-state rates of exit of several molecules from the CSF of unanesthetized goats during ven- triculocisternal perfusion. They found that the permeability of the ventricles in goats (as measured by the KD for the various solutes) decreased in the following order: labelled water (TOH) > urea > creatinine > fructose. This appears to 24 indicate that the flux is inversely related to molecular size. They also observed that the KD for 42K was greater than that for 24Na. Heisey g£_§l. concluded that the passive permeability of the ventricular system in the goat is at least as great as that for the vasopressin-treated toad bladder. Bradbury and Davson (1964) found no evidence for a saturable transport system for either urea or creatinine from the CSF of anesthetized rabbits studied by ventriculo- cisternal perfusion. They found that the flux of urea out of CSF exceeded creatinine efflux,indicating diffusion rate for these two substances is related to molecular size. These workers concluded that there was no carrier-mediated transfer of urea or creatinine, these two molecules exiting the CSF by simple diffusion. Cserr (1965) investigated the potassium exchange 42 between CSF, brain and plasma using K and the technique of ventriculocisternal perfusion in anesthetized dogs and rats. 42K outflux could be recov- 42 She found that two-thirds of the ered from brain tissue indicating that the K in the perfusate was exchanging primarily with brain intracellular potassium 5 pools. Addition of 10' M ouabain to the CSF perfusion fluid 42K outflux to 25 percent of control values, indi- 42 reduced the cating that 75 percent of K outflux is dependent on active ion transport. Cserr ascribed the active step of 42K outflux to cellular components in the brain (and not to the ependyma) for two reasons: (1) large molecules like creatinine or even 25 inulin can diffuse across the ependyma making it unnecessary to assume an active role for ependymal cells; and (2) ouabain 42K~inf1ux. This fact cannot 42 does not inhibit transependymal be explained if the active process for K outflux is placed at the ventricular ependyma. She thus concluded that trans- ependymal potassium exchange is passive. Two other studies on the exchange of potassium between CSF, brain and blood 42 appeared in 1965. Both investigations utilized K and the technique of ventriculocisternal perfusion to measure the flux of potassium between CSF, brain and blood. Bradbury and Davson working with anesthetized rabbits and Katzman 42 et al. using anesthetized cats both reported that K flux from the CSF perfusate did not exhibit saturation kinetics, a finding that is consistent with simple diffusion. Like Cserr, both Bradbury and Davson and Katzman et al. found 42 that ouabain in the perfusate depressed K outflux and ascribed this phenomena to a poisoning of the Na-K pump in cells of the brain parenchyma resulting in an inhibition of potassium influx into cells. Inhibition of potassium flux into brain cells would result in a reduction in the flux of 42K from CSF. Thus both groups of workers concluded that 42K outflux was by a process of simple diffusion. Oppelt et a1. (1963) injected solutions into the 45 cisterna magna of anesthetized dogs containing Ca and either normal or 3-5 times normal concentrations of calcium. 45 They found the Ca removal rate was independent of the 26 injected calcium concentration. These workers concluded there was no active transport of calcium out of the CSF; diffusion and bulk flow accounted for the complete removal of calcium. These results were confirmed by Graziani g£_al. (1965) who found that 45 Ca efflux from the CSF perfusate during ventriculocisternal perfusion in anesthetized cats does not demonstrate saturation kinetics when increasing quantities of unlabelled calcium were added to the perfusate. These workers concluded that calcium efflux from the CSF was 45Ca entering due to passive diffusion with one-third of the brain tissue, the other two-thirds presumably diffusing into blood. Cerebrospinal and Brain Extracellular Fluid’VOlumes As with other tissues, an extracellular fluid (ECS) compartment exists within the central nervous system (CNS). In addition, nervous tissue contains and is surrounded by CSF which, if not the same as brain interstitial fluid, is in communication with it (Tschirgi, 1960). Determination of the volume of these two compartments is helpful for a basic understanding of the chemical environment of the CNS, in the clinical assessment of CSF and CNS abnormalities, in determining intrathecal drug dosages, and in calculating the flux of materials between CSF and brain ECS. 27 Table 2 compares the CSF volumes of various species as determined by several methods; most experiments employ the technique of perfusing the CSF system. Values for total CSF and ventricular volumes are shown,and,so that comparisons can be made among Species, ventricular volume is also ex- pressed as a function of brain weight. Although, among species, there is wide variation in the absolute values for both ventricular CSF and total CSF volumes, the ventricular volume expressed per gm of brain is (excepting the goat) relatively constant. This indicates that, in general, there appears to be a proportionate increase in ventricular volume with increasing brain weight. The data in Table 2 also reveal good agreement between the volumes obtained using the perfusion technique and those determined either by open drainage or ventricular casts. Thus, while similar ventri- cular volumes have been reported for the same species using different techniques (cat and goat, Table 2) the use of dif- ferent methods has caused considerable confusion in the determination of the actual volume of brain extracellular fluid. Over the years a number of methods have been employed, many giving conflicting results. Estimates of the size of the brain ECS have been obtained from electron microsc0pe studies of nervous tissue. These studies have indicated a virtual absence of extracellu- lar space in brain tissue ranging between 0 and 5 percent of the total brain volume (Maynard et al., 1957). These 28 Nama .Emaaooz a aaaaaz mma mmma .uuommEOH w ummq mama ..aa aa casaaaz m m.NN Illll mama ..Hm no aoanncmmmmm mm w.w mama .oaam a maaaom mama ..aa aa aaaam omma ..Hm no oucoaoa mama .aaaxaam a.a ma 0H moa mm mm om mm om NN mummU MNHSUMHH§®> m compSpr mmu :oamSMHQQ o.aa Hmzhopmaooasoaauao> acamzmaom waoqnomamnsm-oH:omau:o> mamas Hmzaoumao Scam owmcamap ooam maoamsmaom amnaoumao-waoczomamnsm w amcaopmaooasoaauco> scam mosam> mo 55m :oamsmaom o.m Hmcaopmaooasowap=o> m.a mummo amasoaauno> :oamsmaom a.a amcamumaooanoahunm> :oamsmama awnE:~-oHsowau:o> :oamsmaom m.o amcaoummooHSUaauco> puswoscm Hmaaoaou m.o anm ommcamaw ovum aaasaaa amen: aaaaaaa amen; moxcoe pmom wow HwU “$0 “$0 oocoaomom maev oESHo> mmu HmpOH mamman sm\ama nHEV pcosoazmmoz mo vogue: masao> amasoaauco> aaeaa< .momoomm oumanopao> pcoaomwap :a moESHo> pHSHm quammOHDoaoo .N aaaaa 29 workers indicated that brain cellular elements were separated by fairly constant gaps, 150-200°A wide. Maynard g£_al, argued that it is unnecessary to postulate the presence of a special blood-brain barrier and suggested that this bar- rier is a lack of ECS in the brain. However, the credibility of these electron microsc0pe studies has declined since the recognition that neural tissue swells considerably and rapidly during the process of fixation (Van Harreveld, 1961; Davson, 1967). Van Harreveld g£_al. (1965) published elec- tron micrographs of rapidly frozen brain tissue, showing enlarged extracellular spaces and yielding values for brain ECS of 18-25 percent of brain weight as compared with approxi- mately 6 percent of tissues that were asphyxiated for 8 minutes before fixation. Allen (1955) incubated brain slices in a medium con- taining inulin or ferrocyanide as extracellular markers (foreign substances that equilibrate rapidly between blood and.interstitialf1uid but do not penetrate cells). He found that the brain spaces occupied by these molecules increased with time presumably as a result of the markers entering cells. By extrapolating back to zero time he estimated a normal brain extracellular space of 14-17 percent. Davson and Spaziani (1959) incubated brain slices with sucrose and chloride. They found that the sucrose space was not affected by incubation time, remaining steady at 13 percent of brain weight. They concluded, however, that their tissue slices 30 were not normal as the chloride-space gradually increased with incubation time,presumably due to the penetration of chloride into cells. Davson (1967) states there are two major sources of error with this in vitro technique of esti- mating brain ECS. First, unless the tissue slice is very thin, oxygenation of the tissue will be poor;causing cere- bral edema,thereby decreasing the size of the ECS. However, cutting the brain tissue slices too thin will cause so many cut ends of cells on the surfaces of the tissue slice that significant penetration of these cut cells by the marker will occur. In addition, any estimate of the brain extra- cellular space obtained by plotting the brain space of the marker against incubation time and extrapolating back to zero time can result in a falsely low value for the ECS size. Davson (1967) indicates that these 13 vitro experiments sug- gest a CNS extracellular space of 10-15 percent but, because the tissue is not normal, they are not convincing. It has been assumed that the CNS extracellular fluid resembles the extracellular fluid of other tissues and that it is the primary container of the two "extracellular ions,” Na+ and Cl' (Tschirgi, 1960). In many tissues the extra- cellular space closely approximates the Na+ and Cl- space. Since the mammalian brain has high concentrations of Na+ and C1-, the size of the Na+ and C1- spaces in brain have been accepted as the size of the CNS extracellular space (Katzman, 1966). Elliot (1949) indicated the rabbit brain had a C1- 31 space of 40 percent of the total brain volume while Bourke g£_al. (1965) found a Cl- space ranging between 20-40 per- cent of brain weight in the guinea pig, rabbit, cat, monkey and sheep. The difficulty with using this method of esti- mating the size of the CNS interstitial Space is knowing the extent to which Na+ and C1. are, indeed, extracellular. Another method of estimating the ECS of brain is to present the extracellular markers via blood (or plasma) to the CNS. Generally the markers used have been inulin, ferrocyanide, sulfate, iodide, thiocyanate, radio-iodinated human serum albumin (RIHSA) and sucrose introduced into the blood by either single intraperitoneal (ip) or intravenous (iv) injection or by constant infusion into blood (Davson and Spaziani, 1959; Reed and Woodbury, 1963; Pollay, 1966; Davson, 1967; Davson and Segal, 1969; Pollay and Kaplan, 1970). This technique has generally given small values for brain space due to the combined effects of their restricted penetration across the blood-brain barrier and the "sink effect" of CSF (Davson and Segal, 1969; Pollay and Kaplan, 1970). Reed and Woodbury (1963) obtained brain spaces in rats of 2 percent for inulin, 6 percent for iodide, and 4 percent for sucrose after giving these markers iv or ip. Davson and Spaziani (1959) found iodide and sucrose spaces in rabbit brains of 3-4 percent following two hours of continuous infusion. Davson and Segal (1969) found a brain sucrose space in the rabbit of about 3 percent following 32 constant iv infusion of this marker for 1.5 hours. However, if the CSF was replaced by silicone (given via a bilateral ven- triculocisternal perfusion) during the iv infusion of sucrose, the sucrose brain space was greater than 6 percent. These workers concluded that CSF acts as a sink and prevents the ex- tracellular fluid from coming into equilibrium with plasma. Removal of the sink action (in this case by perfusing the ven- tricular system with silicone) resulted in a larger measured brain space. Pollay (1966) infused thiocyanate into the ear vein of rabbits and found (at low plasma thiocyanate concen- trations) a brain thiocyanate space of 10 percent of brain weight. If plasma thiocyanate levels were increased (from 1.0 to 3.0 mM/ml) the brain thiocyanate space increased to 20 per- cent. If sodium iodide was given iv or 2,4-dinitrophenol was injected into the CSF, the rabbit brain thiocyanate space was calculated to be 14-15 percent at low plasma thiocyanate levels. Pollay concluded that thiocyanate was actively trans- ported from CSF to blood and this transport was a major com- ponent of the CSF sink action for brain thiocyanate. Altering the effectiveness of this transport process (by saturation, competitive inhibition or inhibiting cellular respiration) reduced the effectiveness of the CSF sink action and gave a more accurate estimate of the brain thiocyanate space. Presentation of the extracellular marker to neutral tissue by way of the CSF has the advantage of circumventing the blood-brain barrier and maintaining the tissue normal. 33 The markers can be introduced either by a single intra- cisternal (IC) injection (Reed and Woodbury, 1963; Bourke e£_al,, 1965; Van Harreveld g£_al., 1966) or by ventriculo- cisternal perfusion (Rall g£_al., 1962; Woodward e£_al., 1967; Baethman g£_al,, 1970; Heisey, 1971). Reed and Wood- bury (1963) found a sucrose brain space in rats of 15-38 percent. Bourke g£_al. (1965) calculated thiocyanate, sucrose and inulin brain spaces for 11 mammalian species. They found the thiocyanate brain space ranged from 17 per- cent in the mouse to 56 percent in the whale. Inulin and sucrose spaces ranged from 8.5 percent to 53 percent in the mouse and whale, respectively. These workers found that the sizes of the indicator Spaces vary among species as a function of the logarithm of the brain weight. Van Harreveld g£_gl. (1966) computed a sulfate space for the rabbit brain of 15-30 percent of the total tissue volume. Rall gt_al, (1962) perfused dog brains and found an ECS (using inulin as an indicator) of 7-14 percent. Woodward g£_al. (1967) found an inulin Space in the rat cerebral cortex of 13.5- 14.5 percent after 6 hours of perfusion. Baethman §£_al. (1970) calculated a thiocyanate space in rat brains of 15.4 percent after 90 min. of perfusion. Heisey (1971) found in turtle brains a RIHSA space of 4-5 percent, an inulin space that varied between 2-10 percent and a fructose space of 14 percent. He indicates that there is an inverse relation- ship between the molecular weight of the extracellular 34 marker and the calculated ECS and that there appears to be a direct relationship between the size of the inulin and RIHSA spaces and the length of perfusion. Longer perfusions apparently allowed time for greater penetration of the inulin and RIHSA from the ventricles into brain, thereby giving larger calculated brain Spaces. Davson (1967) states that unless the marker is completely impermeable to the blood- brain barrier, the blood can act as a sink and drain away the tag presented to the brain by CSF. He emphasizes this point by demonstrating that a sucrose space of 10 percent is found for the rabbit brain when sucrose is presented to the brain by way of the blood and CSF (ventriculocisternal perfusion). When sucrose is presented to the brain by ven- triculocisternal perfusion only, a 6 percent sucrose Space is obtained. The controversy surrounding the size of neural ECS has resulted mainly from electron micrographs of brain tissue that was abnormal due to fixation and because of the slow penetration of molecules from blood to brain. However, with the use of improved tissue fixation techniques for electron micrography and presentation of extracellular markers to the brain by way of the CSF, it appears that the size of the brain ECS is 10-15 percent of total brain weight; a value that is similar to the ECS size in other tissues. 35 Effect of Ionic Changes in Cerebrospinal Fluid on GardIOVascular and Respiratory Functions Many workers have investigated the relationship between the ionic composition of the fluid environment of the CNS and various aspects of CNS function. The majority of these studies have been conducted on anesthetized mammals in which the relationship between changes in the CNS ionic environment and various autonomic functions and/or electri- cal activity of the brain was observed. In general three methods have been used to alter the ionic composition of the fluid environment of the CNS and are classified as follows: a. Intravascular injection of solutions with different ionic composition b. Intraventricular or intracisternal injection of solutions with altered ionic composition c. Perfusion of the CSF system with an ionically altered artificial CSF Most studies have been performed on anesthetized mammals. All anesthetics are known to affect autonomic function (e.g., blood pressure, heart and respiratory rate) either as a result of or in addition to their generalized CNS depressing activity. In general, barbiturates (given intravenously or intraperitoneally) depress respiratory activity and moderately reduce blood pressure in mammals (Goodman and Gilman, 1965). Similarly, administration of 36 the inhalation anesthetics halothane and methoxyflurane is usually accompanied by varying degrees of hypotension and depressed respiration,whereas systemic arterial blood pres- sure generally increases moderately during cycloprOpane inhalation (Emerson and Massion, 1967; Goodman and Gilman, 1965). Anesthesia affects blood pressure in birds as in mammals. Sturkie (1965) indicated that pentobarbital sodium, sodium phenobarbital, sodium barbital and urethane adminis- tered either intravenously or intraperitoneally resulted in a significantly depressed blood pressure (30 mm Hg systolic with pentobarbital sodium). Consequently, in any study of CNS control of cardiovascular function in either anesthe- tized birds or mammals, one must be aware of possible anes- thesia. influences (e.g., attenuation of responses due to experimental alterations) on the results obtained. Hooker (1915) perfused the vascular system of the brains of dogs with blood in which the K+ and Ca++ concentra- tions had been altered. He found that increasing the Ca++ concentration caused a stimulation of the respiratory center (as evidenced by a greater rate and amplitude of respira- tion) and increased heart rate. Conversely, increasing the K+ concentration inhibited the respiratory center and slowed the heart rate. Rubin gt_al. (1943) continuously injected the femoral vein of cats with isotonic solutions of potas- sium chloride or calcium chloride and simultaneously recorded the electroencephalogram (EEG) and electrocardiogram 37 (EKG). They found K+ or Ca++ produced slowing of the EEG at the time of deve10pment of intraventricular block or of cardiac arrest. They concluded that K+ and Ca++ have no demonstrable effect on the EEG of the cat. The results of injecting varying concentrations of potassium and calcium salts either into the brain ventricles or the cisterna magna of mammals are in good agreement. The following studies provide a full survey of the effects of altering the CSF K+ and Ca++ concentrations by intraventri- cular or intracisternal injections in a variety of anesthe- tized mammalian Species: Huggins and Hastings, 1933; Resnik §£_gl., 1936; Mullins g£_al., 1937; Downman and Mackenzie, 1943; Smolik, 1943; Stern, 1945; Walker g£_al., 1945; Cicardo, 1949; Cicardo, 1950; Feldberg and Sherwood, 1957; and C00per et_al,, 1958. In general, when a solution with a high K+-Ca-++ ratio (due either to excess K+ or lowered Ca++) is injected either intraventricularly or intracistern- ally, there is increased muscular activity, agitation, stimulation of reSpiration (greater rate and depth of breathing), a substantial elevation of arterial blood pres- sure and slowing of the heart rate. Slight elevations in the K+ levels sometimes produced an initial fall in arterial pressure along with a bradycardia. Large elevations in CSF K+ (around 10 times normal levels) results in apneusis associated with circulatory collapse. Decreased CSF K+ levels were generally without effect. Increased CSF Ca++ 38 levels resulted generally in a condition of torpor with a decrease in the rate and depth of breathing, a decline in arterial pressure, and a general muscular relaxation. Perfusion of the cerebral ventricles with an arti- ficial CSF containing altered concentrations of K+ and/or Ca++ has confirmed the results obtained with single intra- cisternal injection. Merlis (1940) perfused the lower spinal subarachnoid space in anesthetized dogs with the spinal cord sectioned at T10. The calcium content of the perfusion fluid was altered in these experiments. Merlis found that Ca++-free solutions produced augmentation of the spinal flexion reflex, an increase in muscle tone and spon- taneous twitching of the muscles in the lower half of the body. He attributed the muscle twitching to an increased responsiveness of the spinal cord neurons to normal afferent 11mpulses from the periphery. Leusen (1950) perfused the cerrebral ventricles in anesthetized dogs after bilateral Vagotomy and carotid sinus isolation. He found that an + . . . exx:ess of K in the perfusate caused an increase 1n the artrerial pressure and enhanced vasomotor reflexes. Per- quion with a K+-free solution was without effect. Increased ++ - Ca concentration in the perfusion fluid resulted 1n a 1OWGI‘ing of the blood pressure,whereas lowered Ca” caused a I‘ise in arterial pressure and augmentation of vasomotor refllaaces. Devos (1951) also used a ventriculocisternal perf11$:ion technique in anesthetized dogs. He indicated 39 + . . . that an excess of K 1n the perfusate caused an increase 1n arterial pressure and a reflex bradycardia. Lowered K+ did . 4- not affect blood pressure or heart rate. ngh Ca + concen- trations in the perfusion fluid caused a depression of both arterial pressure and heart rate whereas lowered Ca++ con- centrations provoked a rise in blood pressure and lowered heart rate. Pappenheimer et a1. (1962) found that lowering the Ca++ concentration by 60 percent in the fluid perfusing the cerebral ventricles of unanesthetized goats resulted in the animal becoming restless and excitable. Also evident were large increases in arterial pressure (at times exceed- ing 220 mm Hg) that remained elevated during the perfusion. This hypertension was generally accompanied by a bradycardia and an altered EKG. Horsten and Klopper (1952), using an arrtificial CSF with altered K+ and Ca++ concentrations, Inerfused the cerebral ventricles of anesthetized cats while remzording the cortical EEG. They found increasing the K+ coritent of the perfusion fluid led to formation of spike wavwas in the EEG. K+-free or Ca++-elevated perfusions had no eeffect on the EEG. However, decreasing the Ca++ concen- tratxion of the perfusion fluid resulted in Slow waves (with fast 'waves superimposed) in the EEG. These workers feel that 'the changes seem in the cortical EEG following altera- tHN1S in the ionic composition of CSF were not due to a direczt: action of the ions on the cortex but rather a reflec- tion of an effect on the brain stem. 40 From these studies it appears that the ionic com- position of CSF plays an important role in the maintenance and variation of membrane potentials in the CNS and in the regulation of certain autonomic processes. STATEMENT OF THE PROBLEM The purpose of this study is to determine in the anesthetized chicken, using ventriculocisternal perfusion, the volume, pressure and rates of formation and absorption of CSF; to obtain an estimate of the size of the brain extracellular space; to investigate the movement of radio— iodinated human serum albumin (RIHSA), creatinine, Na, 42K, and 45Ca out of the CSF; and to investigate the effects of changes in the K+ and Ca++ concentration in CSF on cardio- vascular function. 41 METHODS AND MATERIALS A. Animals Sexually mature female single-comb white leghorn chickens (weighing between 1 and 2 kg each were obtained from the Michigan State University Poultry Farm and were maintained at 25 i 1°C with cycles of fourteen hours of light and ten hours of dark. Birds were housed one per cage with feed (MSU cage layer ration) and water given ad libitum. B. Anesthesia l. Phenobarbital sodium Some birds were anesthetized with phenobarbital sodium (170 mg/kg; Merck and Co., Rahway, N.J.) given via the brachial vein. The humerus was exposed and broken to provide an auxillary route for ventilation since there is a unique connection, in birds, between the lungs (via the auxillary air sac) and the humerus air sac. This procedure 'was necessary because chickens anesthetized with pheno- barbital often experience respiratory difficulties due to excessive mucous secretion in the trachea. 42 43 2. 2,2-dichloro-l,l-difluoroethyl methyl ether (Metofane) The remaining animals used in these studies were anesthetized with the inhalation anesthetic, Metofane (Pitman- Moore, Washington Crossing, N.J.). These chickens were brought to surgical stages of anesthesia by placing their heads in a mask supplied with Metofane and compressed air. When an adequate surgical level was obtained (as indicated by pupillary dilation and muscle relaxation), the trachea was exposed, incised, and cannulated with a 3-4 cm length of rubber tubing. This tubing was connected to a T-tube, one arm of which was open to room air and the other to the anes- thesia delivery instrument, providing a non-rebreathing system. Compressed air, the carrier gas, was delivered through a flowmeter (Ohio Medical Products, Cleveland, Ohio) and past a vaporizer (Ohio Medical Products, Cleveland, Ohio) containing liquid Metofane. The amount of Metofane entering the carrier-gas stream was adjusted by means of an adjustable orifice in the vaporizer chamber. Metofane-air mixture was cielivered to the bird at a rate (700-900 cc/min) sufficient ‘to supply the bird's minute ventilation. The amount of hdetofane delivered to the animal could not be determined 'because: (1) With a non-rebreathing system, an unknown amount of Metofane will escape to room air instead of being delivered to the animal; and (2) An unknown quantity of the Metofane delivered to the bird will be sequestered in the 44 various air sacs associated with the avian respiratory system. Pupillary dilation, response to pinching of the comb or toes and general muscle tone were used as qualitative indicators of the stage of anesthesia. C. Surgical Procedures The left femoral artery was cannulated with PE-90 tubing and arterial pressure monitored using a Statham pres- sure transducer (Model P23AC; Grass Instrument Co., Quincy, Mass.) and a Grass polygraph (Model 5; Grass Instrument Co., Quincy, Mass.). The arterial pressure transducer was cali- brated using a mercury manometer; the response to pressure was linear over the range 0-200 mm Hg. With the chicken prone, the head was secured in a stereotaxic frame (Model 4C; H. Neuman and Co., Skokie, 111.), approximately 10 cm above heart level by means of earbars inserted into the external auditory meatus and a pin through the external nares attached to a Y-shaped yoke secured to the support stand of the stereotaxic frame. The animal's neck was acutely flexed so that the parietal surface of the skull assumed an angle approximately 30° from the horizontal. A midline incision from the posterior edge of the comb extended caudally to the fifth or sixth cervical vertebra. A portion of the parietal bone, 7 mm posterior to the bregma and 6 mm lateral to the midline, was abraded with a dental drill and burr (Model 21, Foredom Electric Co., New York, 45 New York) and removed with a small dental spatula, which exposed the dura overlying the left cerebral hemisphere. The atlantooccipital membrane overlying the cisterna magna was exposed by blunt dissection of the complexus and longus colli anterior muscles. The left lateral ventricle was penetrated at a 20° angle to the dura with a short-bevel 27 gauge, 2 inch needle (Vita Needle Co., Needham, Mass.) held in a micromanipulator (Model MM-3; Eric Sobotka, Inc., Farmingdale, New York). The needle was connected to a motor-driven syringe pump (Model 940; Harvard Apparatus Co., Dover, Mass.) and a Statham pressure transducer (Model P23BC; Grass Instrument Co., Quincy, Mass.) by means of a male "T" adaptor and PC-60 tubing (Figure l). A water reservoir (the height of which was adjustable) was used to calibrate the pressure trans- ducer; the response to pressure was linear over the range 0-40 mm H20. Zero hydrostatic pressure was set at the level of the stereotaxic earbars. The perfusion pressure was con- tinuously recorded on a Grass model 5 polygraph. A low flow of artificial CSF was started through the needle prior to its insertion into the brain. Pressure due to the resistance of the inflow needle and tubing was re- corded when the tip of the needle was placed perpendicular to and on the dura mater. AS the needle was lowered slowly through the dura into brain tissue, pressure rose and 46 .cOMpmauaooaoo .0 a onwuso o ”coaumapcoocoo 30amca ..o ”zoamuso :oamswaom .o> ”Boamca coamsmaom .a> “maaswoe .2 menaaonoaoo .u moaozmmaso; Hmapoaoo .mo .mcmme «showmao :H mm oawoo: aoaaoumom moaoaauqo> Hmamuma pmoa a“ ma cacao: Hmmaom .fioamom on azmaw pong paw; sexuago a mo compoom ocaawas Hmufimmmm m ow mono -aomoa :a Eoumxm scamsmaom amasoaauco> mamas mo Emawmam .H oaswam 47 co 0> a a 30:50 llh .umm 292330 memo H oasmam Im40m 1T mmuaomzo 3.32.6 , asst 48 remained elevated. When the ventricle was penetrated (gen- erally 1-2 mm below the dura), there was an abrupt fall in pressure. The cisterna magna was punctured with a short-bevel, 23-gauge, 2 inch hubless needle (Vita Needle Co., Needham, Mass.) fixed in a micromanipulator (Baltimore Instrument Co., Baltimore, Md.) and attached by PE-SO tubing to a reservoir of artificial CSF held 20 cm above the head. The needle, directed anteriorly at a 28° angle to the atlantooccipital membrane was rapidly lowered to a depth of 8 mm below the dura. Simultaneously there was a sharp rise in intraventri- cular pressure; pressure dropped quickly when the tubing was disconnected from the reservoir and the cisternal needle withdrawn 3-4 mm. This fall in pressure concomitant with fluid flow through the cisternal cannula, indicated hydrau- lic continuity through the ventricular system. The cisternal cannula was connected by PE-SO tubing to a photoelectric drop recorder (Model PTTl; Grass Instrument Co., Quincy, Mass.), the height of which could be varied with respect to the stereotaxic earbars (Figure 1). Pressure in the ventricular system was set by the level of the outflow tubing. Intra- ventricular pressure was calculated by subtracting the pressure with the needle tip on the dura from that with the needle tip in the lateral ventricle. 49 D. Cerebral Ventricular Perfusion Technique 1. Perfusion fluid composition In all studies, the total osmolarity of the perfusion fluid was between 280-295 mOsm per Kg H20 and was equili- brated with 4-6% CO2 just prior to use. a. Normal artificial chicken CSF The perfusion fluid used in most experiments was an artificial chicken CSF (Appendix 1) that was similar in composition to normal chicken CSF (Anderson and Hazelwood, 1969). b. Ionically altered artificial chicken CSF In the remaining studies, the perfusion fluid was an artificial chicken CSF with altered Na and K concen- trations (Appendix 1); the other salts remaining in their original concentrations. c. Test molecules added to the perfusion fluids The test molecules used in these experiments were: 1. 22Na (3.0 uc/ml; New England Nuclear Corp., Boston, Mass.) 2. 45Ca (1.0 pc/ml; New England Nuclear Corp., Boston, Mass.) 3. 42K (1.0 pc/ml; Michigan State University Triga reactor) 50 4. carbon-14 labelled glucose (D-glucose-14C; 0.1 uc/ml; New England Nuclear Corp., Boston, Mass.) 5. carbon-l4 labelled creatinine (creatinine carboxyl-14C hydrochloride; 0.1 uc/ml; Nuclear Equipment Chemical Corp., Farming- dale, New York) 6. creatinine (0-2.5 mg/ml; Pfanstiehl Labora- tories Inc., Waukegan, 111.) In all experiments either radioiodinated human serum albumin (RIHSA; 1.0 uc/ml; B.R. Squibb and Sons, Inc., New Brunswick, New Jersey) or inulin (1.0 mg/ml; Pfanstiehl Laboratories Inc., Waukegan, Ill.) was added to the perfusion fluid to measure bulk absorption of CSF. 2. Definition of perfusion period At the start of each experiment the outflow concen- tration of test molecules was zero. Outflow concentrations increased with time with steady-state outflow concentrations reached in 20-30 min (Figure 2). The approach to the steady- state and three 10-20 minute collections of cisternal efflu- ent in the steady-state constituted the first period of an individual experiment. Subsequent periods in the experiment (defined as either a change in intraventricular pressure due to a change in the height of the outflow tubing oraa change in the inflow concentration of K, Ca or creatinine) were 51 .:0amsmaom mo mmuscas om usonm aopmm venomoa mm: paw :Oapmaucoocoo 30amca mo ucooaom aw xaoume -axoumam mm: coaumaucooaoo onmuso opmum->wwoum .HmEazm op oasooaos umou mo compospoapaa moumoawaa oEap oaoN .mmmmaonm ”mouscaev osau coam=maoa mo coauocsm a ma wouHOHQ ma paw moumcapaou :oaumauqooaoo onMGa mo owmucooaom m mm wommoamxo ma omoUSHm-u mo :ofiumauaoocoo Boamuso «H .coxowgo m cm mamas «showmao on“ ow oaoaauco> Hmanoaoo pmoa one anm ucoEHaomxo coamSmhom oawcmm m mo muasmom .N madman 52 N oasmam mN_ ON_~O__ 00. cm mm Om 00 00 tom Om mm ON 9 O. m a.||._|\) _ .4T. _ _|\~7a _ km a _ _ _ _ _ L7 0 $55.2 4 1 1 ON .. 0v J 00 1 Om _ L 00. 8:. 40.. 53 made up of four collections of outflow fluid. The first outflow collection of each period was for at least 30 minutes to assure that steady-state outflow concentration of test molecules 'was reachieved. After the steady-state was attained, three additional 10-20 minute outflow vials were collected and it was from these vials that the outflow rate and test molecule concentration for that period was obtained. 3. Determination of inflow and outflow rates Artificial CSF containing test substances was per- fused into the lateral ventricular needle and collected from the needle in the cisterna magna. Inflow and outflow rates were determined gravimetrically for each experiment. Inflow perfusion rates varied between 38-52 ul/min but were constant for any one experiment. In most experiments inflow rate was determined by collecting fluid from the perfusion syringe over timed periods (10-15 minutes) in tared vials at the beginning and end of each experiment. For the experiments 'where inflow K, Ca and creatinine concentrations were varied, inflow rates were determined at the beginning and end of each period. An average outflow rate for each period was determined from three collections of cisternal effluent (in the steady-state) over timed periods (10-20 min) in tared Vials. 54 4. Determination of test molecule concentration a. In perfusion inflow and outflow The inflow concentration of test molecules was determined in duplicate on 50 ul aliquots from the inflow syringe. Mean outflow concentrations for each period were determined from duplicate or triplicate 50 ul aliquots from the three outflow fluid vials collected in the steady-state 22 42 for that period. Concentrations of RIHSA, Na and K were determined by 1-5 minute counts with a gamma well spectro- meter (Model 530, University II Series; Baird-Atomic, Cam- bridge, Mass.). Carbon-l4 and 45Ca were counted integrally in 10 m1 of scintillation fluid (Aquasol; New England Nuclear Corp., Boston, Mass.) with a trichannel liquid scintillation spectrometer (Model Mark I; Nuclear Chicago Corp., Des Plaines, 111.) (Appendix 2A and 2B, respectively). Non- radioactive K and Ca concentrations were determined by flame photometry (Appendix 3) and atomic absorbance (Appendix 4), respectively. Creatinine was determined colorimetrically using an alkaline picrate method (Appendix 5); inulin by the resorcinol method (Appendix 6). b. In brain tissue Verification of needle placement and the per- fUSion path was made at the end of each experiment by Perfusing methylene blue. The brain was removed, sectioned Sagitally and both halves were examined for staining. The 55 two brain halves were then blotted, weighed, homogenized in water with a teflon pestle homogenizer (A.H. Thomas Co., Philadelphia, Pa.) and diluted to 25 m1. Duplicate 2 m1 aliquots of the homogenate were counted in a gamma well spectrometer for 40 minutes for 131I, 22Na or 42K activity. In addition, duplicate 50 ul samples from the inflow syringe and from two outflow vials (all diluted to 2 ml with water) were counted. To determine 45Ca activity, each brain half was treated as above. Triplicate 5 m1 aliquots of the homo- genates were pipetted into 10 ml of Aquasol scintillation fluid forming an opaque, stiff gel. These were counted integrally in a trichannel liquid scintillation spectrometer (Appendix 2B). E. Calculations The equations to be presented have been derived by Heisey et a1. (1962). Sample calculations are shown in Appendix 7. 1. Definition of symbols V rate of flow, ul/min 1,0 = subscripts referring to inflow and outflow, respectively f,a = subscripts referring to formation and absorp- tion of CSF, respectively 56 c concentration, quantity per unit volume C clearance of RIHSA or inulin, ul/min B 2. Rate of CSF formation, Vf The fluid balance in the ventriculocisternal per- filsion system is assumed to be (Literature review, p. 9). V + V. = V + V (1) Rearranging (1) gives the equation used to calculate (38F formation. Vf = Va + (Vo - Vi) (2) V0 and Vi are measured gravimetrically (Methods, D-3). Thus, in order to calculate Vf, Va must be determined. 3. Rate of CSF bulk absorption, Va Evidence will be presented that loss of RIHSA and inulin from the avian ventricular system is predominantly by bulk absorption of CSF distal to the fourth ventricle (i.e., diffusion of these two molecules from the ventricular system is negligible). The calculation of Va is then v=c=i° ° (3) 4. Clearance of smaller molecules from the ventricular system For molecules smaller than RIHSA or inulin which can leave the ventricular system by simple diffusion or active transport, clearance is calculated by: 57 i i 0 CK = - ° (4) c where CX = clearance of a small molecule x, ul/min a = mean ventricular concentration of x The calculation of E was defined by Pappenheimer et a1. (1961) as c = C0 + 0.37 (ci - co) (5) 5. Outflux coefficient, KD (non-bulk clearance) The clearance of molecules smaller than RIHSA or inulin from the CSF system is made up of two components: (1) clearance due to bulk absorption; and (2) clearance due to diffusion and/or active transport. The clearance of the small molecule by bulk absorption is given by: C c B (43-) (6) C X where: co = ratio of the cisternal outflow and _E—. x mean ventricular concentrations of the small molecule, x. Subtracting the bulk clearance of x from its total clearance gives clearance of x due to diffusion or active transport, i.e., K (7) where: = non- bulk clearance of x, ul/min 58 KD is also called the outflux coefficient of x. Expanding (7) gives the equation used to calculate KD. B o (8) 6. Outflux rate, n The outflux rate of x from CSF by diffusion or active transport is calculated as the difference between the rate at which x enters the ventricular system by perfusion (Vici) and the rate at which x leaves the CSF in the cisternal out- flow (Voco) and by bulk absorption (CB - co). Stated mathe- matically: n = KD ' c = Vic. - V c - (CB - c ) (9) where: fix = outflux rate of x, quantity per minute 7. Mass balance determination In order to further quantify the amount of material leaving the ventricles and to determine its final disposition (i.e., brain parenchyma or blood), a mass balance for four 22 45 42 radioactive molecules (RIHSA, Na, Ca and K) was deter- mined using the following equations. a. Total radioactive input The total amount of radioactivity (IT) added to the system from the perfusion inflow is equal to the inflow 59 rate (Vi) times the inflow concentration (ci) times the total perfusion time (t). Stated mathematically: 1T = (vici) t (10) b. Total radioactivity collected in perfusion effluent The total amount of radioactivity collected from the cisternal effluent (OT) is given by the perfusion outflow rate (V0) times the concentration in individual effluent samples (co) times the collection time of each sample (At) summed for all samples collected during an experiment (n). Stated mathematically: (voco) At (11) 0 ll 0M3 c. Total radioactivity cleared from the ventricular system The total radioactivity cleared from the ventri- cular system (CT) is given as the product of the total CSF clearance (Cx) and mean ventricular concentration (Ex) of molecule x times the clearance time of each period (At) summed for all clearance periods in an experiment (n). Stated mathematically: CT: OMB (cxéx) At (12) 60 d. Total radioactivity cleared from the ventricular system by bulk absorption The total radioactivity removed from the ventri- cular system by bulk absorption (FT) is given as the product of the clearance of RIHSA or inulin (CB) and the outflow concentration of molecule x (Cox) times the clearance time for each period (At) summed for all clearance periods (n). Stated mathematically: n FT = 3 (CB c ) At (13) e. Total radioactivity in brain tissue The total radioactivity in brain tissue (BT) was determined by multiplying the radioactive concentration of x in an aliquot of homogenized brain (cxb) times the dilu- tion factor for the brain homogenate (r) times the total brain weight (WB). Stated mathematically: B = (cx - T b r) WB (14) The calculations for the mass balance determina- tions were performed on the Michigan State University CDC 6500 computer (Appendix 8). 8. Distribution volume The distribution volume (which is an estimate of the ventricular volume) of each molecule was determined by 61 integrating the outflow concentration approach to the steady- state with respect to time (Figure 2). Distribution volume was calculated using the following equation: n O O V 3 EVicit - Voc0t - CxctilAt D ' - ’ Vdsci (15) c where: VD = distribution volume of x, ul c. ,c = concentrations of x in the inflow l1: °t and outflow, respectively at time t ct = mean concentration of x in CSF during time At (Equation 5) c = mean ventricular concentration after distribution of x is com- plete Vds = volume of the perfusion needles and tubing (dead space volume), ul n = number of samples 9. Brain space calculation Assuming that the concentration of radioactive mate- rial in the brain wasiJlequilibrium with the mean concentra- tion in the ventricular system in the steady-state, brain space was calculated as: 62 be - 100 Sx = RCSF (16) where: 8x = brain space of x, percent of brain weight be = brain radioactivity, cpm/gm W I CSF - CSF radioactivity, cpm/ml F. Cardiovascular Measurements Blood pressure was recorded on the Grass polygraph and heart rates were taken from the blood pressure records. Mean blood pressure was calculated from the following equa- tion (R.K. Ringer, personal communication). mean pressure = Z (d1asto%1c pressure) + systol1c pressure (17) Blood pressures and heart rates were measured for 30 minutes prior to and after perfusion with an unaltered artificial CSF had been initiated. The values obtained were used as control measurements, each animal serving as its own control. Blood pressure and heart rate was deter- ‘mined at three separate intervals during the steady-state portion of the experimental perfusion period and the results were expressed as a percent of control values. 63 G. Statistical Analysis All data obtained were statistically analyzed using either continuous simple linear regression; student's "t" (paired or group comparisons) or analysis of variance (Sokal and Rohlf, 1969). A probability of 0.05 was used as the level of significance in all statistical tests. RESULTS Cerebrospinal Fluid Formation and Absorption‘Rates The cerebral ventricular system of the chicken was perfused with an artificial chicken CSF containing either RIHSA or inulin. The clearance of either of these molecules was used to estimate the rate of CSF bulk absorption (Va; Equation 3, Methods) and CSF formation rate (Vf; Equation 2, Methods). Figure 3 illustrates the relationship of CSF formation, RIHSA clearance (CRIHSA) and perfusion outflow minus inflow rates (VG-Vi) to intraventricular pressure (P). Data used to calculate lines in Figure 3 were obtained from 19 steady-state perfusion periods in 6 animals. Intraven- tricular pressure was altered by varying the height of the outflow cannula; each change in perfusion pressure repre- sented a different clearance period (section D2, Methods). C increases linearly and (VG-Vi) decreases linearly RIHSA with intraventricular pressure over a range from 0 to 20 cm H20. Absolute values for the SIOpeS of Vo-Vi) on P (-0.26 ul/min-cm) and CRIHSA on P (0.22 ul/m1n-cm) are not stat1st1- cally different (P > 0.05), which is interpreted to indicate that RIHSA is removed from CSF predominantly by bulk absorp- tion thereby justifying its use in the chicken to measure 64 65 H mm.a + m amm.m “a a~.m- .>-o> am.o + m amm.m “a NN.m u "mam cam :oammoawoa panama mo wocuos 0;» an woumasoamu who: momma How mcoaumscm . mmmaom m “Omm Eu mm whammoam am soaapco>mauca o :oauoc: a mm . a a . m . m pouuoam noumcawao gamma m.:aE\H1 ma>-o>v moumh BoHMGa pew 30am -pso amasowauco> coozuon ounoaommfiw pew .noumaawao puma mnws\az mv mama coaumsaom mmo new: .oHSmmoam amasuaauqo>maunw mo coawozsm m mm mouma 30amca cam BoamuSo awasoaaaco> noozuon monoaomme paw .oocmamoao 0.05) over the range 0-20 cm H20, the implication being that intraventricular pressure had no effect on CSF formation rate in chickens. Where the ‘ regression lines for CSF formation and absorption are super- imposed, the point of intersection (Figure 3; arrow) repre- sents the normal rate of formation (i.e., where formation and absorption are equal) and the normal physi010gica1 pressure in the CSF system. A value for intraventricular pressure (4 cm H20) determined in this way coincides with that of CSF pressure measured directly from the cisterna magna in 20 experiments (4.2 i 0.3 cm H20). Mean rate of CSF formation (calculated from 56 steady-state perfusion periods in 21 animals) was 1.4 i 0.1 ul/min. Molecular Flux from Cerebrospinal Fluid to Brain and Blood The artificial CSF used in these studies contained 22 42 45 (in addition to RIHSA or inulin) Na, K, Ca or creatinine in order to evaluate the permeability characteristics of the avian cerebral ventricular system. Evidence given previously (Figure 3) indicates that the ependymal linings of the chicken ventricular system are essentially impermeable to large molecules like RIHSA and inulin and that these mole- cules are removed from CSF by a pressure-dependent bulk flow 22 process. For smaller molecules (like Na and creatinine) my data indicate that they are also removed from CSF by a 68 pressure-dependent process but, in addition, they exit by simple diffusion and/or active transport. Data presented in Figure 4 (from 14 steady-state perfusion periods in five 22 animals) illustrate the separation of total Na clearance (CNa; Figure 4B; Equation 4, Methods) into bulk absorption (Figure 4C; Equation 6, Methods) and non-bulk clearance (Figure 4A; Equations 7 and 8, Methods) components. The 22Na absorbed in bulk with RIHSA or inulin is the outflux coefficient (KD ) for 22Na Na (Figure 4A). The regression line relating KD to P is not Na Significantly different from zero (P > 0.05), indicating difference between CNa and the that over the pressure range 0-20 cm H20, KD is independent Na of intraventricular pressure. This shows the pressure- sensitive part of CNa to be bulk absorption. Non-bulk clear- 22 ance accounts for most of the Na lost from the CSF system at low intraventricular pressures. At higher pressures, the 22 amount of Na lost by both routes is approximately the same. Data in Figure 5 (from 16 steady-state perfusion periods in five animals) Show the division of total crea- tinine clearance (Ccr; Figure 5B) into bulk clearance (Va; Figure 5C) and non-bulk clearance (KD ; Figure 5A) compon- cr ents. The regression line of KD vs P is not different cr from.zero (P > 0.05) indicating that, like KD- , KD is Na cr independent of intraventricular pressure over the range 0.20 cm.HZO. At low intraventricular pressures, the amount 0f 0.05) indicating that over the [cr]i range 0-225 mg/100 ml, KDcr is not affected by [cr]i. A concentration-independent KD is consistent with but does not conclusively prove simple diffusion. Creatinine outflux rate (ncr) as a function of inflow creatinine concentration ([cr]i) is shown in Figure 7. Over Over the concentration range 0-225 mg/100 ml, her increases linearly with [cr]i illustrating a concentration dependent 74 ao m.aa + aaaua aam.m aa aam.m- n ma ”ma use :oammoamoa amocaa mo voguoe one an woumHDOHmo mm: mama on» now :oapmSGm a one .AHE ooa\me ”.maoHV :oapmaucoocoo ocacapmoao onmnH no .mmmaomn< .n.:ae\an a QMV ucoaoammooo Knamuso ocanapmoao .oumcapao .coaumaucoocoo onapmoao zoawca coamsmaom mo coauocsw m mm wouuon uqoaoamwooo xzamuso ocacapmoau .o oasmHm 75 9a flaw .3 wk 02 m8 0 oHDMam oo. n~ mu .1 0. cu 76 a . I . ao m-aa x ma.a + .aaum amm m +a a a u a ”ma mam scammmamma ammqaa an mmumasoamo mmz mnaa msa aom :oaumscm mae .AHE coa\me mamaogv coaumaanmocoo mcacaammao 30amaa .mmmaOmn< .m.¢aE\mE mao my mama xzamazo maacaammao .mamcamao .aoaamaaamocoo mcanaammao 30Hmca scamsmamm mo soauocsm m mm mmauoam mama Knamaso mcacaummao .m mammam 77 m- .31 2.. 0m. mu. m masmfim co. 1 2. 50 Yo: c on ow om ow. on. ow. EN 03 cam 78 flux of creatinine from CSF. That ncr is not saturable at high [cr]i implies either creatinine efflux from CSF is by non-carrier mediated diffusion, the carrier has not been saturated, or the carrier is not located at the ventricular ependyma. Figure 8 is a plot of 42K outflux coefficient (KD ) K as a function of the inflow potassium concentration ([K]i) in the perfusion fluid for 16 steady-state perfusion periods in six animals. The regression line is not different from 42 zero (P > 0.05) suggesting that K exits the CSF system by passive diffusion at least over the [K]i range 0-31 mEq/L. Further indication of the passive 42K outflux from 42 CSF is shown in Figure 9 which is a plot of the K outflux rate (nK) versus-the inflow potassium concentration ([K]i)° The 8K increases proportionately to [K]i revealing no satura- 42 tion of K outflux from the avian ventricular system up to 31 mEq/L. 45 The Ca outflux coefficient (KD ) and calcium out- , Ca flux rate (nCa) as a function of inflow calcium concentration ([Ca]i) for 8 steady-state perfusion periods in three animals appear in Figures 10 and 11, respectively. The regression line relating K and [Ca]i (Figure 10) is not significantly D Ca different from zero indicating that over the concentration range 0-5 mEq/L, KD is independent of [Ca]i. 45Ca flux Ca from CSF is a positive linear function of [Ca]i (Figure 11). 79 an H m.aa + .aa_ amm.m “a ma.m- u a ”ma mam mameazm a ma mmoaamm mommammao mumam-%©mmum ca aom :Oammmamma ammcaa mo wogame map >@ pmamasoamo mm: mqaa ma» aom coapmscm mgb .mmmmaomnm mq\cme mammav coaumaacmocoo enammmaom seawca .mmumaamao ”caE\H: ”gnaw acmaoammmoo azamaso Edammmaom .coaumaacmoaoo Enammmaom 30awca coamsmamm mamam> acmauammmoo xzamuso Edammmuom .w maswam 80 w maswam 0. .ON on Figure 9. 81 Potassium outflux rate as a function of perfusion inflow potassium concentration. Potassium outflux rate (nK; uM/min.; ordinate). Inflow potassium concentration ([K]i; mM/L; abscissa). The equation for the line was calculated by the method of linear regression and is: 8K = 0.92 (1 0.12) [K]i + 3.3 x 10‘2 40' 30 20 D “O- 82 20 Figure 9 813 Figure 10. 83 Calcium outflux coefficient plotted as a function of perfusion inflow calcium concentration. Ordinate. Calcium outflux coefficient (KDca, ul/min.). Abscissa. Inflow calcium concentra- tion ([Ca]i, mEq/L). Equation for the line was calculated by the method of linear regression and is: KD = -0.04 (t 0.4) [Ca]. + 4.6 Ca 1 IO 84 Figure 10 Figure 11. 85 Calcium outflux rate as a function of perfusion inflow calcium concentration. Ordinate. Calcium outflux rate (nCa, qu/min.). Abscissa. Inflow calcium concentration ([Ca]i; mEq/L). Equation for the line was calculated by linear regression and is: - -4 nCa = 56.8 c: 1.5) [Ca]i - 0.39 x 10 300 250 200 I50 IOO 1 U C -4 axIO 86 2 3 Figure 11 87 45 The data are interpreted to indicate that Ca efflux from the ventricular system is by passive diffusion. The effect of perfusion time on the outflux coeffi- 22 45 42 cients for Na, Ca, K and creatinine are summarized in Table 3. Data in this table reveal that extended perfusion 22 45 42 time has no effect on the KD's of Na, Ca, K or crea- tinine and indicates that passive permeability of the ventri- cular ependyma was not altered by perfusion. Table 3. Effect of perfusion time on the efflux coefficients for sodium, calcium, potassium and creatinine. Perfusion Outflux Coefficients (ul/min) Time (hrs.) Sodium Calcium Potassium Creatinine 0-2 4.0 1 0.8 5.3 i 1.1 15.6 i 1.4 12.4 1 2.2 (6) (3) (6) (14) 2-4 4.0 i 0.5 3.0 13.0 i 1.8 13.2 i 2.0 (5) (2) (6) (11) 4-6 4.6 i 0.4 4.6 1 0.5 13.1 i 1.7 14.3 i 2.6 (5) (3) (5) (6) Note: Values are means i SEM. Number of observations given in parentheses. Data in Table 3 Show that there is no difference (P > 0.05) between KD and KD or between KD and KD but Na Ca K cr the latter two KD'S are approximately three times greater than those for sodium and calcium. This shows that the 88 chicken cerebral ventricles are more permeable to potassium and creatinine than to either sodium or calcium. Mass Balance Determination In an attempt to determine the extent to which test molecules enter either blood or brain tissue from CSF, the total amount of radioactivity presented to the animal, the total amount recovered in the perfusion outflow and the total amount cleared from CSF were calculated in addition to meas- uring the total amount of radioactivity recovered from brain tissue (Methods, Section D4b). Table 4 is a radioactivity balance sheet for RIHSA, 22Na, 45Ca and 42K. Values are expressed as a percent of either total input (Equation 10, Methods) or total cleared from CSF (Equation 12, Methods). Rows l and 2 are total radioactivity collected in the per- fusion effluent (Equation 11, Methods) and total radioactiv- ity cleared from CSF (Equation 12, Methods), respectively, and both are expressed as a percent of total input. It is assumed that all radioactivity presented to the animal by perfusion inflow will either be collected in the cisternal effluent or will be cleared from CSF into brain or blood. Thus, the sum of row 1 and 2 for each molecule should equal 100 percent; that they do not reflects errors associated with measurement of inflow and outflow rates and concentra- tions and errors in the estimation of clearance. Approxi- mately 97 percent of the radioactivity (for all test 89 .AaaaUaaaaa .me a :mamm macaw mHmEHcm mo amnesz :mmE mm wmmmmamxm mmwmanmoamm ampoz wmammHo Hmuoa m.~ a m.aa m.am N.“ a m.- m.~ a a.mH Ho m mm caman :a Hmuoa m . I . . . I . . I . “saga amaoa mo a a H + a mm o m m o + m m m o + m o- mm mmammHo HHsnIeo: Hmaoa m . I . . . I . . I . gamma amaoa mo m m o + m m a w m H + m m o H + m m mm HHsn :H mmammHo Hmuoe m smna m o a.a a a.am a.ma a.a a a.aa a.a a a.a mo a ma aawaaam #mwow a . I . . . I . . I . wanna Hmuou mo H mm acmsHmmm Ham Haa ama Hma a aaaa>ouam m Mme 0mm mzNN Haomonmm .Mmm mam .mUma .mZNN . .v mHnma 90 molecules) introduced at the perfusion inflow could be accounted for in that collected in effluent fluid and that cleared indicating a 3 percent measurement error. The total clearance of molecules from the CSF is made up of a bulk and non-bulk clearance. Total bulk clear- ance (Equation 13, Methods) expressed as a percent of total input appears in row 3 (Table 4). Non-bulk clearance (expressed as a percent of total input, row 4) was calcu- lated as the difference between total clearance and bulk clearance and the sum of values in rows 3 and 4 for each molecule Should equal the value in row 2. The data Show 22 45 that approximately 50 percent of the Na and Ca cleared from CSF is by bulk absorption whereas only about 12 percent of the total 42K cleared is by this route; most 42K exiting by non-bulk means. There is no difference (P > 0.05) in total RIHSA clearance and that absorbed in bulk, providing additional justification for using RIHSA clearance to meas- ure CSF bulk absorption in the chicken. The negative non- bulk RIHSA clearance (Row 4) is probably due to errors in measurement of inflow and outflow rates and concentrations. Row 5 (Table 4) shows the uptake of RIHSA, 22Na, 45Ca and 42K by brain tissue as a percent of the total clear- ance from the perfusion fluid. The results show that brain uptake accounts for approximately 50 percent of 42Kiclear- 45 22 ance, 27 percent of Ca clearance and 23 percent of Na clearance from CSF, the remainder presumably going into 91 42 blood. While K flux out of CSF appears to be divided equally between brain and blood, approximately three-fourths d 45Ca efflux is into blood. Approximately 16 of 22Na an percent of the total RIHSA clearance is into brain tissue. It is interesting to speculate how this occurs since the non-bulk clearance of RIHSA has been shown to be essentially zero (Figure 3; Table 4, row 4). The data suggest that RIHSA is cleared into blood from the ventricular system by bulk absorption but then in some manner enters brain tissue from blood. Cerebrospinal Fluid Volumes and Brain Spaces The cerebral ventricular volume was estimated from 22 14C-glucose the volumes of distribution of RIHSA, Na and as shown in Figure 12. The calculation of the volume of distribution of any substance depends upon the amount of the substance remaining in CSF after a steady-state concentra- tion is attained, taking into account the amount of the substance cleared from CSF and the amount in the perfusion tubing (Equation 15, Methods). There is no difference (P > 0.05) in the three distribution volumes (approximately 140 pl), indicating that this volume is independent of the test molecules' weight and size. Total CSF volume in the chicken is approximately 350 pl and is estimated from the maximum volume that can be aspirated from the cisterna magna. 92 22 Figure 12. Mean distribution volumes of RIHSA, Na and 14C-glucose in cerebral ventricular system of the chicken (ul, ordinate). The number of animals used is shown in parenthesis and SEM is designated by the vertical lines. 93 rDISTRIBUTION VOLUME (H) II IIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIII I IIIIII II I I I IIIIIIIIIIIII ’DDDD >7: IIIII I I I I I IIII III II IIII III I I I I IIIIIIIIII IIIIIIIIIII I IIIIIIII IIIIIIIIIIIIIIIIIIIIII II IIIIIIII II I IIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIII I IIIIIIIIIIIIIIIIIIIIIIIIII I III IIIIIIIII IIIIIIIIII IIIIIIIIII IIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIOIIIIIII IIIIIIIIIIIIIII IIIomIwan-IIIIIIb-III“-IDPDPIIDIIIIIDIDbbrbbblPIIIIDIIDDIDDIDDDPPDFDPPIDDDIDDDDD I0. I III II I II”IIII.IIIHIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII dldu IImuudnu-uuuuuddlIII-IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII v IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII. III IIIIIIIIIIIIIIIIIIIIIIIII IIIIIO IIIIIIIIIIIIIIIIIIIIIIIII IIIII I I I I v I I II I v. I v I v. I I I III "I III I IIIIIIII I I I IIIIIIIIIIIIIIIIIII.IIII I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII I II III-IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII II I II I IIIIIII IIIIII II I III IIIIIIIIIIIIIIII IIIIIIIIIIIIIIIII IIIIII‘IIIIIIIIIII IIIIII. IIIIIIIIIIII (Io) IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII I IIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIOIIIIO. IIIIIIIIIIIIIIIIIIII IIIII III I AAA’AuAAAA RIHSA IIIfI '4C-Glucose I50? IHI I—I )- Figure 12 94 Data reported in Table 5 show that brain spaces ranged from 4 percent of brain weight (RIHSA) to 91 percent of brain weight for 42K. Brain spaces determined with dif- ferent molecules were different (P < 0.05) from each other. In general, there appears to be an inverse relationship between the molecular size of the test molecule and the calculated brain space, however,this alone cannot fully explain the four different brain spaces obtained. Table 5. Distribution of RIHSA, 22Na, 45Ca, and 42K in the chicken brain. Brain Space* Test Molecule Number of Animals (% of brain weight) RIHSA 10 4.0 i 0.3 22Na 7 10.8 i 1.0 45Ca 3 27.5 i 3.8 42K 6 91.3 i 10.7 *Mean i SEM. 95 Effects of Altered Cerebrospinal Fluid Potassium and EaICium Concentrations on cardiovascular Functions The role of CSF ionic composition in cardiovascular function has been extensively investigated in anesthetized mammals. In order to study this in the chicken, the cere- bral ventricles were perfused with solutions containing altered concentrations of potassium and calcium. I first had to determine if ventricular perfusion affected cardio- vascular parameters. Data in Table 6 illustrate the effect of ventriculocisternal perfusion on blood pressure and heart rate in chickens anesthetized with either sodium phenobarbital or Metofane. Blood pressure (in mm Hg) are shown as systolic over diastolic and as the mean (Equation 17, Methods). Sys- tolic, diastolic and mean blood pressure and heart rate in animals anesthetized with sodium phenobarbital were lower (P < 0.05) than those anesthetized with Metofane both before and after ventricular perfusion had begun. Ventricular per- fusion caused an increase (P < 0.05) in the systolic, dias- tolic, and mean blood pressure in sodium phenobarbital- anesthetized chickens but did not alter heart rate (P > 0.05). In animals anesthetized with Metofane, ventricular perfusion produced no significant change (P > 0.05) in blood pressure or in heart rate. In Figure 13 systolic (panel A), diastolic (panel B), and mean (panel C) blood pressure responses (expressed as a percent of control values; Section F, Methods) are plotted 96 .momonuaohmm cw czocm mHmEficm mo honezz .2mm H mamas mm commoumxo mozfiw> "muoz head fiHNV fiHNU mafiv flHNV mHNU I H mm a w om mammOpoz A H “mm m H HHH m w NIH A A New I w Isa m H NmH mmfiv fimfiv flmfiv ammo nmNV mNNU m H «w I H mo Hmufinuaa -oconm m w QNN m H mm I H mHH m H ONN a.am m H IoH asfieom some uwaoummfiw name oflHoummflw ncfia\mumonv ufiaoumxm mqfle\muwmnv owHOpmzm mama ovum Uwpmgummé “Imam flmm may ppmom flmm may onsmmohm poon whammonm pooam mhnuufifim HNHDUHHHQO> mhspudfim HNHSUMHHQoxr houm< whomom .mnoxuflgu a“ mum“ Home: cam whammopm pecan so :ofimsmhom HmH30Mhpco> Hahnohou can mwmonumonm mo uummmm .o canoe 97 Figure 13. Experimental systolic, diastolic and mean blood pressure (expressed as a percent of control blood pressure) plotted as a function of perfusion in- flow potassium concentration. Ordinate. EXperi- mental systolic (A), diastolic (B), and mean (C) blood pressures (all expressed as a percent of control blood pressure). Abscissa. Inflow potas- sium concentration ([K]i; mEq/L). Equations for the lines were calculated by the method of linear regression and are: % of control (systolic) = 0.44 (i 0.2) [K]i + 92.3 % of control (diastolic) = 0.60 (i 0.3) [K]i + 88.2 % of control (mean) = 0.43 (i 0.3) [K]i + 92.8 98 PERCENT OF CONTROL BLOOD PRESSURE ”OFA IZO " 80b 1 1 n i ‘4 80 .. I I . I O 60 .. I; L 1 L .1 MO- (3 0 I20 ’ . C C O . C C IOO - ° F ' I I [K]. L. L I l 1 _J 0 IO 20 30 40 Figure 13 99 as a function of inflow potassium concentration ([K]i). Data are from 18 steady-state periods in eight animals anesthe- tized with Metofane. Although all three regression lines have positive slopes (suggesting an increase in blood pres- sure with an increasing [K]i)’ none are significantly different from zero. These data indicate that CSF potassium concentrations in the range 0-40 mEq/L did not affect blood pressure in anesthetized chickens. In Figure 14, experimental systolic (panel A), dias- tolic (panel B) and mean (panel C) blood pressures (expressed as a percent of control pressures) are plotted as a function of inflow calcium concentration ([Ca]i) for 11 steady-state perfusion periods in five Metofane-anesthetized animals. All three regression lines have similar negative slopes; however, only the one for systolic blood pressure (Figure 14, panel A) is different from zero (P < 0.05). The data sug- gest that altering CSF calcium concentration affected blood pressure but variability in the responses did not permit statistical significance to be achieved for diastolic and mean blood pressure. 100 Figure 14. Experimental systolic, diastolic, and mean blood pressure (expressed as percent of control blood pressure) versus perfusion inflow calcium concen- tration.. Ordinate. Experimental systolic (A), diastolic (B), and mean (C) blood pressures (all expressed as percent of control blood pressure). Abscissa. Inflow calcium concentration ([Ca]i; mEq/L). Equations for lines were calculated by the method of linear regression and are: % of control (systolic) = -4.4 (i 1.9) [Ca]i + 97.0 % of control (diastolic) = -4.9 (i 2.6) [Ca]i + 98.6 % of control (mean) = -4.7 (t 2.3) [Ca]i + 98.0 101 PERCENT or CONTROL BLOOD PRESSURE mo W0 80 H0 [00 90 70 mo I00 80 60 1 T F Figure 14 DISCUSSION Cerebrospinal Fluid Formation and Absorption Cerebrospinal fluid (CSF) formation has been measured either by draining CSF from the cisterna magna over a measured time period or by measuring the time required to restore CSF pressure to pre-withdrawal levels after removal of a known volume of fluid. The major weakness of such methods is that at normal intraventricular pressures, not all of the fluid formed will be collected; an unknown amount will be lost to blood by bulk absorption. CSF formation, determined by free drainage techniques, was determined at low intraventricular pressures and CSF loss by bulk absorption was assumed to be negligible. However, low intraventricular pressures do not guarantee the complete cessation of bulk absorption. Validity of the values for CSF formation obtained using free drainage techniques are questionable. The technique of ventriculocisternal perfusion has provided the most precise and, probably,the most physiologi- cal method of studying CSF formation and absorption. One advantage of this method is that bulk CSF absorption can be estimated. Thus, normal CSF formation and absorption rates (at physi010gica1 intraventricular pressure) can be 102 103 determined in addition to studying the effects of altering intraventricular pressure on both parameters. In this study on chickens as in previous studies on one reptilian species (Heisey and Michael, 1971) and various mammalian species (Heisey et a1., 1962; Bering and Sato, 1963; Cutler gt_al., 1968; Katzman and Hussy, 1970), the rate of CSF bulk absorption was linearly related to intra- ventricular pressure (Figure 3). However, the CSF bulk absorption rate in chickens is less than reported for mam- mals (Cutler §t_gl,, 1968) and turtles (Heisey and Michael, 1971). Resistance to bulk absorption (the reciprocal of the slope of CRIHSA plotted as a function of P (Cutler g£_al,, 1968) in the chicken (4,545 cmomin/ml) was 12-350 times the resistance reported for mammals (Cutler g£_al., 1968) and 2.5 times that reported for turtles (Heisey and Michael, 1971). These latter workers suggested that the large resist— ance to bulk absorption in the turtle might be caused by the absence of an arachnoid membrane with associated "valves." However, Hansen-Pruss (1923) has demonstrated that birds possess an arachnoid membrane. The high resistance in chickens may indicate either a lack of valve-like structures in the arachnoid villi, as described for mammals (Welsh and Friedman, 1960), or presence of high resistance pathways in the arachnoid membrane. CSF formation rate in the chicken was found to be huflependent of intraventricular pressure (Figure 3) 104 paralleling results previously reported for the goat (Heisey et_213, 1962), cat (Hochwald and Wallenstein, 1967; Katzman and Hussy, 1970), dog (Bering and Sato, 1963), human (Cutler g£_al., 1968) and turtle (Heisey and Michael, 1971). The mean value of CSF secretion in 21 chickens was 1.4 ul/min. Total choroid plexus weight was 3.6 i 0.4 mg in 7 chickens. Thus, formation rate per unit weight of choroid plexus was 0.39 ul/min per mg tissue, a value that is similar to values obtained in other Species (Table 1). In addition, the CSF turnover rate in the chicken was 0.4 percent/minute, a value which is comparable to those calculated for other species (Table 1) and which reveals that the ventricular CSF volume (140 pl) is replaced every 1.7 hours and that the total CSF volume (350 pl) is replenished every 4 hours. Heisey g£_al, (1962) estimated that choroid plexus blood flow in the goat would be about 3 ul/min per mg tissue to get a CSF secretion rate of 0.3 pl/min per mg tissue assuming 20 percent of the plasma flow through the choroid plexuses was lost forming CSF. Welch (1963) measured a choroid plexus blood flow of 2.86 ul/min per mg tissue, a blood volume loss of 13.3 percent during transit through the plexuses and calculated from these values a CSF forma- tion rate of 0.37 ul/min per mg tissue for the rabbit. In the chicken, for a CSF formation rate of 0.39 pl/min per mg tissue (assuming that, like the rabbit, 13.3 percent of the blood flow through the choroid plexuses is secreted as 105 CSF) choroid plexus blood flow would be 2.9 ul/min per mg tissue, a value which is essentially the same as that reported for the rabbit and goat. Choroid plexus blood flow in all three animals is about 5 times greater than to whole brain and 1.4 times less than to the kidney, indi- cating that choroidal tissue has a comparatively large blood flow. Molecular Flux from Cerebrospinal Fluid Tb Brain and Blood‘* Development of the technique of ventriculocisternal perfusion has provided a useful method for the quantitative study of the transport of substances among blood, brain and CSF. The technique has been used in this study to investi- 42 45Ca from CSF to gate the movement of creatinine, K and brain and blood. All test molecules investigated had a clearance from the perfusion fluid which was greater than inulin or RIHSA, indicating that, in addition to their loss from CSF by bulk absorption, these molecules left the ven- tricular system by simple diffusion and/or active transport. Before molecular flux of materials from the perfusion fluid can be studied, it is necessary to determine what effects perfusion has on the integrity of the ventricular walls. Cserr (1965) has shown, in dogs, a doubling of the creatinine outflux coefficient after 4-6 hours of perfusion. She indicated that this increase in creatinine outflux was 106 due to a progressive increase in ventricular permeability resulting from deterioration of the preparation with time. There was no change in KD , KD , KD , and KD with time Na Ca K cr (Table 3) suggesting that cerebral ventricular perfusion (of up to 6 hours) did not alter the passive permeability of the chicken ventricles. Permeability of porous membranes to inert, lipid- insoluble molecules varies inversely with molecular size. If any lipid-insoluble substance is cleared from the ven- tricular system at a rate greater than that predicted for. diffusion based on molecular size, this may be taken as presumptive evidence that the molecule is actively trans- ported although further experimental evidence is required to prove the point conclusively (Heisey g£_gl,, 1962). These workers found that, in goats, the outflux coefficient of creatinine (MW=113) was less than that of smaller mole- cules and was greater than that of fructose (MW=160). Bering and Sato (1963) found that the urea outflux coeffi- cient in the dog was three times greater than that for creatinine. Cserr (1965) reported for both rats and dogs, 22 42K efflux coefficients which exceeded that of Na and creatinine. Results from all three studies indicate that creatinine movement from mammalian CSF is by passive dif- fusion. These results are contrary to data of Bierer (1972) who found that, in dogs, the creatinine outflux coefficient exceeded that of para-aminohippuric acid (PAH), a molecule 107 which he demonstrated to be actively transported from the ventricular system. He suggested that creatinine may be actively tranSported from CSF. Ventricular permeability to creatinine, in the 42K and is approximately three times larger than that for 45Ca or 22Na. This sug- chicken, is the same as that for gests either that some mechanism retards the movement of the ions from CSF or that creatinine may be actively trans- ported from the avian ventricular system. If 42K, 22Na and 45Ca movement from CSF is against an electrical gradient, their movement could be impeded so that the outflux coeffi- cient of a non-electrolyte (like creatinine) might appear large by comparison. However, Held §£_al. (1964) and Bradbury and Btulcova (1969) demonstrated that in mammals CSF was electropositive with respect to blood, providing an electric field that would accelerate the passage of these cations from CSF. Assuming that, like mammals, the CSF of chickens is electrically positive with respect to blood (although it has never been measured) would suggest that the large creatinine outflux coefficient cannot be explained by an electrical gradient hindering ionic outflux. In the present study, the creatinine outflux coef- ficient was independent of creatinine concentration in CSF (Figure 6) and creatinine outflux was a linear function of CSF creatinine concentrations (Figure 7) suggesting (but not conclusively proving) that creatinine apparently exits 108 CSF by passive diffusion. This indicates that creatinine outflux is not limited by competition for sites on a carrier at the concentrations (--20 mM/L) used in this study. These results confirm those of Bradbury and Davson (1964) who perfused the ventricular system of rabbits with "low" (1.7 mM/L) and "high" (34 mM/L) concentrations of creatinine. They found no evidence of a saturable system involved in creatinine movement from CSF and concluded that non-carrier mediated diffusion described creatinine movement. The large creatinine outflux coefficient (relative to that for 42K, 22 45Ca) might suggest either an active transport Na, and process that does not involve carrier mediation, a carrier which is saturated at creatinine concentrations above 20 mM/L, or the carrier is not located at the ventricular ependyma. Transport of creatinine from the kidney tubules of chickens has been postulated by Shannon (1938), Sykes (1960) and Rennick (1967) who reported that the ratio of creatinine clearance to inulin clearance was greater than one but approached unity with increasing plasma creatinine concentrations. In addition, Rennick demonstrated that this renal tubular transport could be competitively inhibited. Further experimentation with metabolic and competitive inhibitors will be required to determine if creatinine movement from chicken CSF involves active transport. Data presented in Table 3 show that ventricular 42 22 13ermeability to K is 3-4 times that for Na. The lack 109 42 of change in the K outflux coefficient with increasing CSF potassium levels (Figure 8) coupled with the dependence of 42K outflux on concentration (Figure 9) suggests that removal 42 42 of K is by simple diffusion. Passive movement of K out of CSF has previously been demonstrated in various mammalian species (Bradbury and Davson, 1965; Bradbury and gtulcova, 1969; Katzman, 1965; and Cserr, 1965), and in addition, 42 Heisey et a1. (1962) and Cserr (1965) found that K outflux always exceeded that for 22Na by approximately four times. 42 It is likely that the large K outflux observed in chickens and mammals results from brain cells acting as a 42 "sink" for K. Because of the high concentration of potas- sium in brain cells relative to that in interstitial fluid or CSF (Bradbury and §tulcova, 1970), there is probably a 42 large capacity for K exchange with the unlabelled intra- cellular potassium. Data in Table 4 show that 50 percent 42 22 Na leaving CSF enters of the K and 23 percent of the brain tissue, confirming findings in the dog (Cserr, 1965) and cat (Katzman et a1., 1965). This intracerebral potas- 42 sium pool presumably acts as a "sink" for K and is one 42 factor that is responsible for the large K outflux coef- ficient (relative to 22Na outflux coefficient) observed. In addition, molecular size may contribute to the difference in ventricular permeability to 22Na and 42K. When ions are in solution, water molecules become intimately associated 'with the ion resulting in solvation or hydration 110 (Dowben, 1971). Ions with the smallest radii have the largest and most strongly held layer of water molecules. The sodium ion has a smaller radius than the potassium ion, but with its shell of water molecules, its diameter is 1.2 times larger than that of the potassium ion. Sodium ions should diffuse from the ventricles more slowly than potas- 22 sium which would contribute to the difference between Na 42 and K outflux coefficients. Data presented in Figures 10 and 11 demonstrate that, 42K, 45Ca movement from chicken CSF does like creatinine and not exhibit saturation kinetics and confirms findings in dogs (Oppelt et al., 1963) and cats (Graziani et al., 1965). 45Ca efflux is by diffusion unless a This suggests that carrier for calcium is unsaturated at twice the normal cal- cium concentration in CSF. Cerebrospinal Fluid Volume and Brain Spaces The avian brain contains bilateral cerebral and optic lobe ventricles, a third and fourth ventricle, and a small cerebellar ventricle. While most mammalian ventricles can be described as well-formed invaginations (more or less centrally located in their respective brain areas), chicken ventricles are little more than narrow slits. The cerebral ventricles are located superficially under the posterior surface of the cerebral hemispheres as evidenced by their penetration at a point 1-2 mm below the dura. The 111 ventricular volume of chickens averaged 140 pl which is less than half the total CSF volume (350 pl) estimated for these animals (Results). Total ventricular volume (expressed per unit weight of brain tissue) was 45 ul/gm of brain and falls within the range of values reported for several mammalian species (Table 2) indicating that even though the avian brain has a greater number of CSF cavities than the mammalian brain, it contains a similar volume of ventricular CSF. The wide variance in brain spaces calculated for 42 22Na and 45 RIHSA, K, Ca (Table 7) is probably due to: (l) differences in ventricular permeability to the test substances; (2) the extent to which the different molecules enter brain tissue or blood from CSF; and (3) the degree to which the molecules enter cells. Heisey (1971) reported a RIHSA brain space in turtles that was 3-5 percent of brain weight and which varied directly with perfusion time. This range of values encompasses the RIHSA brain space calculated for chickens (Table 5). How- ever, in chickens, there was no relationship between calcu- lated RIHSA space and perfusion time. The small RIHSA space is probably due to the low permeability of the ventricular ependyma to RIHSA (Figure 3, Table 6). It has been suggested previously that the rapid 42K movement from CSF occurs because brain cells, with high intracellular potassium concentration, act as a "sink" for 42 42 K. Consequently, K molecules entering brain tissue 112 would exchange with this intracellular potassium. Combined 42K from CSF into brain tissue 42 with the large flux of (Table 4), distribution of K molecules into the intra- cellular space of brain is apparently responsible for the 42 large K brain space. Tschirgi (1960) assumed that sodium is contained primarily in the extracellular fluid of brain tissue and reported a 30-35 percent sodium brain space in mammals, a value substantially larger than the sodium brain space (11 percent) measured in the chicken (Table 5). Data in Table 4 indicate that only about 20 percent of the 22Na leaving CSF is recovered in brain tissue; presumably most 22 of the Na enters blood. Perhaps blood acts as a "sink” for 22Na leaving CSF and continually drains it away from brain parenchyma. The superficial location of the lateral ventricles within the cerebral hemispheres may result in long diffusional pathways to the bulk of brain tissue and 22 could cause incomplete distribution of Na in brain ECF. Both mechanisms would result in an underestimation of the 22Na brain space. 45Ca brain space determination is subjected to the same source of error associated with incomplete distribution between ventricular fluid and brain extracellular fluid as 22 45 22 Na, yet the Ca brain space exceeds that of Na by 2.5 times (Table 5). Data in Tables 3 and 4 reveal that there is no difference in ventricular ependymal permeability to 113 22Na and 45Ca and the flux of these two molecules from CSF into brain tissue is the same. Giese (1968) has reported that calcium can readily bind with intra- and extracellular 45Ca brain 45C proteins. This suggests that the size of the space (relative to that for 22Na) may be influenced by a binding to proteins. ' Effects of Altered Cerebrospinal Fluid Potassium and Calcium Concentrations on Cardiovascular Functions Effects of CSF potassium and calcium concentrations on the mammalian cardiovascular system has been investigated by several authors and has been extensively reviewed by Tschirgi (1960) and Winterstein (1961). A marked rise (up to 180 mm Hg above control) in systemic blood pressure con- comitant with a reflex slowing of the heart is the response to either raising the potassium concentration or reducing the calcium concentration in CSF. Raising the CSF calcium concentration above normal values usually causes a signifi- cant decline in blood pressure whereas lowering the CSF potassium levels was without effect on blood pressure. These cardiovascular responses to changes in CSF composition are considered to be due to a direct action of these ions on the medullary vasomotor center (Tschirgi, 1960; Winterstein, 1961). The blood pressures, following changes in the CSF potassium concentrations (Figure 13), were not different 114 from control values (P > 0.05) indicating a lack of influ- ence of CSF potassium on the blood pressure of chickens. Systolic blood pressure was depressed (P < 0.05) by increased CSF calcium concentrations (Figure 14, Panel A) but no changes in diastolic and mean blood pressure were detected (P > 0.05). With the exception of the systolic blood pres- sure response to altered CSF calcium levels, these results on chickens do not confirm previous findings in mammals. Absence of a blood pressure response in birds could occur because: (1) the avian brain does not possess a vasomotor center controlling blood pressure; (2) changes in CSF ionic composition were not reflected in the interstitial fluid surrounding the vasomotor center; (3) the avian vasomotor center is normally insensitive to changes in the ionic com- position of CSF; or (4) the anesthetic used in these studies reduced the responsiveness of the vasomotor center to altera- tions in CSF ionic composition. It is difficult to deter- mine which of the above reasons might be responsible for the lack of cardiovascular response in birds to changes in the ionic composition of CSF. Systolic blood pressure measured from a cannulated carotid artery in unanesthetized, restrained adult white leghorn hens averaged 160 mm Hg with a pulse pressure of 25 mm Hg while mean heart rate in these animals was 357 beats per minute (Sturkie, 1965). These blood pressure values may underestimate "normal" values as restraint per se 115 has been shown to reduce blood pressure in white leghorn hens (Whittow §t_al., 1965). Sturkie's values are consider- ably higher than those presented in Table 6 for chickens anesthetized with either sodium phenobarbital or Metofane. Sodium phenobarbital lowered blood pressure more so than Metofane suggesting that Metofane does not depress central nervous system activity as much as sodium phenobarbital. fl A Since chickens anesthetized with Metofane had blood pressures and heart rates that more closely approximated unanesthetized values, all the cardiovascular studies were performed on éJ chickens anesthetized with Metofane in an attempt to mini- 9 mize any masking effect by anesthesia of cardiovascular responses to changes in CSF ionic composition. Sturkie (1965) presumes that, like mammals, the avian brain has a vasomotor center located in the medulla although he states that no conclusive experiments on this have been conducted. Rodbard and Tolpin (1947) found a positive relationship between body temperature and blood pressure in the chicken and inferred from previous studies on turtles (Rodbard, 1947) that neural control of blood pressure was present in chickens. Birds are known to possess a "neural pool" located in the medulla that is responsible for the central control of respiration (Marshal, 1960; Sturkie, 1965). Sturkie (1965) indicates that anesthetics can either stimulate or inhibit this center depending on the anesthetic used. The reduction of blood pressure and 116 heart rate in the anesthetized chickens of the present study (Table 6) compared with unanesthetized chickens (Sturkie, 1965) might be interpreted as additional evidence suggesting some central control of blood pressure and heart rate. Recently, Cohen e£_al. (1970) using the technique of retro- grade degeneration, discovered that the cell bodies of vagal cardio-inhibitory fibers in the pigeon reside in the dorsal motor nucleus. Electrical stimulation of this nucleus resulted in bradycardia, a response that was abolished by bilateral vagotomy (Cohen and Schnall, 1970). Since there appears to be central control of heart rate and respiration in birds, it seems likely that the brains of these animals would possess the capability for control of blood pressure. It is impossible to determine if changes in the CSF ionic composition were reproduced in the interstitial fluid surrounding the vasomotor center. The fact that both 42K 45Ca had measurable fluxes into brain and appreciable and brain spaces suggests that calcium and potassium entered brain tissue. If the vasomotor center was in proximity to the path of the perfusion fluid, presumably changes in CSF ionic composition were reflected in the extracellular fluid around this center. It appears unlikely that the avian vasomotor center is normally insensitive to changes in the CSF ionic composi- tion as increased calcium concentrations in the perfusion fluid caused a moderate but significant decline in systolic 117 'blood pressure (Figure 14, Panel A). The inability to statistically detect responses in diastolic and mean blood pressure (Figure 14, Panels B and C, respectively) to changes in the CSF calcium levels or in systolic, diastolic and mean blood pressure (Figure 13, Panels A, B and C, respectively) following alteration in the CSF potassium concentration is ‘probably due to the variability of the response as indicated by the scatter of points around the regression lines in Figures 13 and 14. Similarly, in pilot studies on chickens anesthetized with sodium phenobarbital, the blood pressure and heart rate responses to changes in the CSF potassium and calcium levels were also variable and inconsistent. .Downman (1943) reported that in rabbits anesthetized with either urethane or sodium phenobarbital, intracisternal injection of potassium phosphate increased blood pressure in some cases and depressed it in others, even in the same animal. He attributed this variability of response to the ‘use of anesthetics in their experiments. Mullins g£_al, (1938) found, in dogs, that increases in blood pressure :following intracisternal injection of calcium-free solutions <>r solutions with elevated potassium concentrations were Egreater in unanesthetized than in barbital or sodium pento- 1>arbital anesthetized animals. In Rodbard and Tolpin's (1947) study in chickens, the use of anesthetics (sodium Iflmenobarbital) was discontinued because it tended to attenu- ate; changes in blood pressure that resulted from changes in 118 loody temperature. The inability to show changes in blood jpressure as a result of CSF ionic alternations is probably (due to the variability of the blood pressure response that Inay have resulted from anesthetic interference with the sensitivity of the vasomotor center. Perhaps chickens that xvere deeply anesthetized did not respond to the same degree as more lightly anesthetized animals. To answer this ques- tion conclusively, studies with unanesthetized animals or animals where the circulating anesthetic concentrations can ‘be monitored and adjusted are required. SUMMARY Chicken brain ventricles were perfused with an artifi- cial CSF yielding data on CSF formation and absorption rates, molecular movement from the CSF, size of the ventricular and brain extracellular Spaces, and effects of CSF ionic changes on cardiovascular functions. Clearance of RIHSA and inulin from CSF was a positive linear function of intraventricular pressure and was a measure of CSF bulk absorption. CSF formation rate was 1.4 i 0.1 ul/min and was inde- pendent of intraventricular pressure from 0-20 cm H20. KDcr and KDNa were independent of intraventricular pressure from O to 20 cm H20. KDNa’ KDCa’ KDK and KDcr were unaffected by perfusion time indicating perfusion did not alter the permeability of the brain ventricular walls. KDNa and KDCa were of the same order of magnitude as were KDK and KDcr’ however the latter two KD's exceeded those for sodium and calcium by three times indicating that chicken cerebral ventricles were more permeable to potassium and creatinine than to either sodium or calcium. 119 1CL 11. 12. 120 Kpcr, KDK and KDCa were each independent of their perfu- 51on inflow concentration while ncr’ nK and nCa were posi- tive linear functions of inflow perfusion concentration. The mechanism of creatinine efflux was not determined. Creatinine outflux coefficient (KD) was independent of concentration which implies diffusion whereas the large KD for creatinine (relative to KDNa’ KDCa’ and KDK) is presumptive evidence for active transport. 42 45Ca outflux did not exhibit saturation That K and kinetics was interpreted to indicate that these ions leave CSF by simple diffusion. Chicken brain ventricular volume was approximately 140 pl and total CSF volume was estimated to be 350 pl. Brain spaces (expressed as a percent of brain weight) 22Na, 45Ca, and 42 for RIHSA, K were 4 percent, 10.8 percent, 27.5 percent, and 91.3 percent, respectively. Size of the calculated brain spaces depended on: differences in the ventricular permeability to the test molecules; the extent to which the different molecules enter brain tissue or blood; and the degree to which the molecules enter brain cells. With the exception of a significant (P < 0.05) decline in systolic blood pressure as a result of increasing CSF calcium levels, variability of the blood pressure response prevented any statistical detection of blood pressure changes that occurred following alterations 121 in the CSF calcium or potassium concentrations in anes- thetized chickens. This suggests that chickens are normally capable of responding to variations in the ionic composition of CSF, but perhaps anesthetics attenuated the blood pressure response. APPENDICES APPENDIX I PREPARATION OF ARTIFICIAL CEREBROSPINAL FLUID 1\rtifici Na+ K+ Ca++ Mg++ Ileagents 1. 43-01 APPENDIX I PREPARATION OF ARTIFICIAL CEREBROSPINAL FLUID al chicken CSF contains: Cations Anions 155 mEq/L Cl- 140 mEq/L 3.7 mEq/L HCOS- 23 mEq/L 2.5 mEq/L 2.1 mEq/L Total osmolality = 287 mOsm/Kg H20 NaCl, analytical reagent (A.R.) KCl, A.R. NaHCO A.R. 3, CaCl A.R. 2’ MgCl 6 H o, A.R. 2 ' 2 Dextrose (Fisher Scientific Co., Fairlawn, N.J.). S tock Solutions 1. Normal CSF Dissolve 7.72 gm NaCl, 0.2759 gm KCl and 1.93 gm NaHCO3 in distilled water; q.s. 1 liter. 122 123 Ionically altered CSF Dissolve 4.77 gms NaCl and 1.93 gm NaHCO3 in dis- tilled water; q.s. 1 liter. MgCl2 Dissolve 21.35 gms MgCl 6 H20 in distilled water; 2 q.s. 100 ml. CaCl2 a. Dissolve 13.87 gms CaCl2 in distilled water; q.s. 100 ml. b. Dissolve 11.10 gms CaCl in distilled water; 2 q.s. 100 m1. ZPerfusion fluids A. Artificial chicken CSF (normal composition) One-tenth m1 of solutions #3 and #4a and 125 mg dextrose was added to 100 ml of solution #1. Artificial chicken CSF (altered calcium concentration) Perfusion fluid containing 0-5 mEq/L Ca was made by adding 0 to 0.025 ml of solution #4b, 0.01 ml of solution #3 and 12.5 mg dextrose to 10 ml of solu- tion #1. Artificial chicken CSF (altered potassium concentra- tion) Changing the potassium concentration of the perfusion fluid between 0 and 40 mEq/L would result in the perfusion fluid becoming either hypo- or hyperosmotic. 124 When KCl was added in excess of normal concentration (3.7 mEq/L), an equal amount of NaCl was removed to maintain isotonicity. Solution #2 was potassium- free and contained 40 mEq/L less NaCl than that in normal CSF perfusion fluid (solution #1). Crystal- line KCl and NaCl were added to solution #2 so that their combined contribution was 40 mEq/L. This was accomplished by adding 0-29.8 mg KCl (0-40 mEq/L K+) 0-23.4 mg NaCl, 12.5 mg dextrose and 0.01 ml of solutions #3 and #4 to 10 ml of solution #2. The osmolality of this fluid was 280-295 mOsm/kg H 0. 2 APPENDIX II LIQUID SCINTILLATION COUNTING APPENDIX II LIQUID SCINTILLATION COUNTING Reference: Instruction Manual, Mark I liquid scintillation counter Model 6860, Nuclear Chicago Corp., DesPlaines, I11. Principle: Liquid scintillation counting is a method of detect- ing radioactivity by means of a solution of fluors (liquid scintillator) and a photomultiplier tube. The energy emitted by the radioactive material is converted to light energy by the liquid scintillator which is then detected by the photo- multiplier tube connected to amplifiers and a scaler circuit. The radioactive sample is placed close to the fluor by dis- solving, immersing, or suspending it in the liquid scintilla- tor. Consequently, this technique is particularly well-suited for use with low energy beta emitters such as tritium and carbon-l4. A. Carbon-14 detection In these studies carbon-l4 was counted alone. In two experiments it was determined that the quenching of all samples was equal. The counts per minute (cpm) and not the cutlculated disintegration rate (dpm) was used to determine Carbon- 14 concentrations . 125 126 B. Calcium-45 detection Calcium-45 activity was measured in both the per- 45 fusion fluid and brain tissue. The quenching of Ca activity was approximately five times greater with brain 45 tissue than with perfusion fluid samples. To compare Ca activity in both perfusion fluid and brain, it was necessary 45Ca activity in terms film} 45 to eliminate quenching effects and get of its actual disintegration rate (dpm). Ca standards, in both the perfusion fluid and in brain tissue were prepared and 45Ca counting efficiency determined in both media. £4 Knowing the efficiency, the disintegration rate of 45Ca I activity in the unknown samples was calculated from the counts per minute. 1. Preparation of calcium-45 standards a. Perfusion fluid standards. Five standards were prepared by adding 0.05 uc 45Ca in 0.05 ml of artificial (111,000 dpm) of chicken CSF (perfusion fluid A, Appendix 1) to 10 ml of scintillation fluid (Aquasol; New England Nuclear Corp., Boston, Mass.) b. Brain tissue standards Five standards were prepared by adding 0.05 uc (111,000 dpm) of 45Ca in 5 ml of homogenized chicken brain tissue (Methods, part D4b) to 10 ml of Aquasol forming a stiff, opaque gel. 127 Counting procedure and calculations The efficiency is calculated using the equation: _ NS . 100 where: ce = efficiency of calcium-45 counting (%) Ns = net count rate of calcium-45 standards (Cpm) Ds = known disintegration rate of calcium- 45 standards (dpm) The disintegration rate of all unknown samples was determined using the equation: D it; u Ce disintegration rate of unknown samples (dpm) net count rate of unknown samples (6pm) where: Du 2 ll u APPENDIX III FLAME PHOTOMETRY APPENDIX III FLAME PHOTOMETRY Reference: Instruction Manual, Model 105 flame photometer, Beckman Instruments Inc., Fullerton, California Principle: The electrons of certain neutral atoms (such as sodium and potassium) can be raised from the ground state to an excited state by heat from pr0pane-oxygen flame. In returning to the ground state, the excited electrons emit radiation of a specific wave-length which is characteristic for each atom. The intensity of the radiation is propor- tional to the number of atoms excited. Detection of the emitted radiation is by a phototube which is connected to amplifiers and a readout meter. Reagents: l. Lithium concentrate (18 mg/ml, Li+; Beckman Instru- ments Inc., Fullerton, Calif.) 2. Potassium standard solutions (0,5,10 mEq/L K+; Beckman Instruments Inc., Fullerton, Calif.) 3. KCl, A.R. 128 129 Solutions: A. Lithium working solution Mix 5 ml of reagent 1 with deionized water; q.s. 1 liter. This solution is used to dilute all standards and unknown samples. B. Potassium working standards 1. Add 0.5 ml each of potassium (0,5 and 10 mEq/L) standard solutions (0,5 and 10 mEq/L) to solu- tion A; q.s. 100 ml. These are 1/200 dilutions of 0,5 and 10 mEq/L potassium standards, respec- tively. 2. Two additional potassium standards (20 and 40 mEq/L) were prepared by adding 0.1492 and 0.2984 gm KCl to solution A; q.s. 100 ml. The reproducibility of these standards was used to verify the commercial standards (B-l). Instrument‘Standardization: 1. Zero mEq/L potassium working standard (B-l) was aspirated into the atomizer for 5 seconds and the instrument adjusted to read zero mEq/L potassium. The procedure was repeated in triplicate. 2. The remaining potassium working standards (B-1; 5 and 10 mEq/L) were used to set the output span of the instrument and to determine the linearity of the response. 130 Procedure: Duplicates of unknown samples were diluted 1/200 with solution A.* Samples containing between 11-20 mEq/L potassium along with the 20 mEq/L potassium standard were diluted an additional Z-fold (final dilution, 1/400) while samples with 21-40 mEq/L potassium (and the 40 mEq/L potas- sium standard) were diluted an additional 4-fold (final dilution, 1/800) with solution A. All samples were aSpirated into the atomizer for 5 seconds and their potassium concen- trations (in mEq/L) were determined from the standards. APPENDIX IV ATOMIC ABSORPTION SPECTROSCOPY APPENDIX IV ATOMIC ABSORPTION SPECTROSCOPY Reference: Instruction Manual, Model 290B Atomic Absorbance Spectrophotometer, Perkin-Elmer Co., Norwalk, Conn. Principle: Atomic absorption spectrosc0py is based on the principle that neutral atoms of certain elements (such as calcium and magnesium) can absorb energy. The energy is supplied from a hollow cathode lamp, the cathode of which is constructed of the element under study. Energy emitted by the excited atoms from the cathode is characteristic of the metal of which the cathode is made. As a result, a calcium cathode will emit energy of a wavelength that only calcium atoms can absorb. The function of the flame in an atomic absorption instrument is to isolate the neutral, ground state atoms in the sample. The amount of radiation absorbed is detected by a phototube and is proportional to the number of atoms present in the sample. Reagents: 1. Lanthanum oxide, LaZO3 (American Potash and Chemical Corp., West Chicago, 111.) 131 132 2. Brook standard calcium solution (1 mg/ml CaCOS; Aloe Scientific, St. Louis, Mo.) 3. CaCl A.R. 2’ Solutions: A. Lanthanum stock solution Mix 58.64 gm La203 in 50 ml of deionized water. Slowly add 250 ml 0.8N HCl to dissolve the La 0 . 2 3 Dilute to 1 liter with deionized water. B. Lanthanum working solution ' Dilute 30 ml of solution A with deionized water; q.s. 1 liter. This solution is used for diluting all standards and unknown samples. C. Calcium working standards 1. Add 0.5, 1.0 and 2 ml of the calcium standard solution (reagent 2) to solution B; q.s. 200 ml. These are 2.5, 5 and 10 mEq/L calcium standards, respectively. Solution B was used as the zero calcium standard. 2. Two additional 5 and 10 mEq/L calcium standards were prepared by adding 0.02775 and 0.0555 gm of CaCl2 to solution B q.s. 100 ml.' The repro- ducibility of these standards was used to verify the commercial standards (C-l). 133 Instrument Standardization: 1. Solution B was aspirated through the nebulizer and the instrument was zeroed. The procedure was repeated in triplicate. 2. The other calcium working standards (C-l; 2.5, 5 and 10 mEq/L) were used to set the output span of the instrument and to determine the linearity of the response. Procedure: 1 Duplicates of unknown samples were diluted 1/200 with solution B, aspirated for 20 seconds through the nebu- lizer and their calcium concentrations (in mEq/L) were determined from the standards. APPENDIX V CREATININE ASSAY 1"“; '1 APPENDIX V CREATININE ASSAY Modified from S. Natelson, 1961 Principle: Picric acid forms a colored complex in alkaline solution with creatinine, with maximum absorbancy at 490 mu. Color intensity is pr0portional to the concentration of creatinine. In addition, color intensity is dependent on the temperature, concentration of alkali and picric acid and time, but these factors were maintained constant in the analysis. Reagents: 1. Picric acid (J.T. Baker Chemical Co., Phillipsburg, New Jersey) 2. NaOH, A.R. 3. HCl 4. Creatinine (Pfanstiehl Laboratories Inc., Waukegan, Ill.) Solutions: A. Picric acid (1.0%) Dissolve 10.0 gm picric acid in distilled water; q.s. 1 liter 134 135 B. NaOH (10%) Dissolve 100.0 gm NaOH in distilled water; q.s. 1 liter C. HCl (0.1 N) Dilute 8.5 ml conc. HCl in distilled water; q.s. 1 liter Creatinine standard solutions: Creatinine is placed in a desiccator over calcium chloride or heated at 100°C in an oven to remove moisture. 200 mg of the dried creatinine is dissolved in 0.1 N HCl h q.s. 100 ml (2.0 mg/ml creatinine). Dilute 8.75, 7.50, 6.25, 5.00, 3.75, 2.50 and 1.25 ml of 2.0 mg/ml creatinine solu- tion to lO‘ml with 0.1 N HCl obtaining 1.75, 1.50, 1.25, 1.00, 0.75, 0.50 and 0.25 mg/ml creatinine standards, respec- tively. Standards are layered with toluene, stoppered and stored at 4°C to prevent their deterioration. Procedure: Mix 8 parts of solution A with 2 parts of solution B forming alkaline picrate. Allow mixture to stand for 10 minutes before use. To duplicate 0.025 ml water blanks, standards and unknown samples add 2.5 ml alkaline picrate and mix in pyrex test tubes. After 10 minutes add 2.5 ml distilled water and mix. Read optical density (0.D.) of all samples against the water blank at 490 mu within 136 15 minutes in a Spectrophotometer (Model DB; Beckman Instru- ments, Inc., Fullerton, Calif.). Calculations: Optical density at 490 mu plotted as a function of creatinine concentration in the standards yields a straight line over the range 0-2.0 mg/ml creatinine. Creatinine concentration in the unknown samples is calculated by multiplying the optical density of the unknown samples by the lepe of the standard curve. CS Cu=6ra XODu concentration where: C OD Optical density 5 = standard u = unknown a = average APPENDIX VI INULIN ASSAY APPENDIX VI INULIN ASSAY Direct Resorcinal Method Without Alkali Treatment. Modified from H. W. Smith, 1956, p. 209. Principle: Inulin is hydrolyzed to fructose by heating in acid and the fructose molecules combine with resorcinol to yield a colored complex. The intensity of the color is propor- tional to the amount of fructose present. Maximum absorbancy of the complex is 490 mu. Reagents: 1. Resorcinol (Fisher Sci. Co., Fairlawn, New Jersey) 2. Ethanol (95%) 3. HCl 4. Inulin (Pfanstiehl Laboratories, Inc., Waukegan, I11.) SOlutions: .A. Resorcinol (1.0 mg/ml) Dissolve 100 mg resorcinol in 95% alcohol; q.s. 100 ml. 137 138 B. HCl (approximately 10 N) Add 776 ml of concentrated HCl to 224 ml of distilled water Inulin standard solutions: Dissolve 200 mg of inulin in distilled water; q.s. to 100 ml (2.0 mg/ml). Dilute 7.5, 5.0, 4.0, 3.0, 2.0, and 1.0 ml of 2.0 mg/ml inulin solution to 10 ml with distilled water obtaining 1.5, 1.0, 0.8, 0.6, 0.4 and 0.2 mg/ml inulin standards, respectively. Storage and satbility of solutions: StOpper all standards and store in refrigerator at O-4°C. Resorcinol (solution A) is prepared fresh daily. HCl (solution B) is stable indefinitely at room temperature. Procedures: To duplicate 0.05 ml water blanks, inulin standards and unknown samples, add 1.0 ml solution A and 2.5 ml solu- tion B and mix under a hood in pyrex test tubes. A glass marble is placed on the top of each tube and the tubes are incubated for 25 minutes at 80°C in a water bath. The tubes are cooled to room temperature and the Optical density at 490 mu determined against the water blank within one hour in a spectrOphotometer (Model DB; Beckman Instruments, Inc., Fullerton, Calif.). 139 Calculations: Optical density at 490 mu plotted as a function of inulin concentration in the standards yields a straight line over the range 0-2.0 mg/ml inulin. Inulin concentration in unknown samples is calculated by multiplying the optical density of the unknown samples by the Slope of the standard C s C = -——- x OD u [ODé] a u where: C = concentration curve . OD = Optical density 5 = standard u = unknown a = average APPENDIX VII SAMPLE CALCULATIONS OF CEREBROSPINAL FLUID FORMATION AND ABSORPTION RATES, OUTFLUX COEFFICIENT, AND OUTFLUX RATE APPENDIX VII SAMPLE CALCULATIONS OF CEREBROSPINAL FLUID FORMATION AND ABSORPTION RATES, OUTFLUX COEFFICIENT, AND OUTFLUX RATE Data in Table 7-1 are measured quantities from a ven- triculocisternal perfusion experiment which allow calculation of CSF formation and bulk absorption rates, CSF clearance and outflux coefficients for sodium and creatinine, and flux rate of creatinine. Table 7-1. Primary data for chicken 3c. Perfusion Perfusion Flow Rates Inflow and outflow concentrations 22 . . RIHSA Na Creatinine Vi Vo Ci Co Ci Co Ci Co ul/min ul/min CPm/ul Cpm/ul Cpm/ul Cpm/ul ug/ul ug/ul 38 37 161.8 156.6 92.4 82.3 2.21 1.79 140 141 Calculations from primary data 1. CSF bulk absorption (Equation 3, Methods) 0 V.C. - V C V = C i i o a RIHSA co 0 (38)(16l.8% - £37)(156.6) = 2.3 ul/min 56 2. CSF formation (Equation 2, Methods) Vf = Va + (Vo — Vi) = 2.3 + (37-38) = 1.3 ul/min 3. 22Na calculations a. Mean ventricular concentration (Equation 5, Methods) C C0 + 0.37 (Ci-co)==(82.3) +0.37 (92.4-82.3) 86.0 fifig. b. 22Na clearance (Equation 4, Methods) Na (.2 c. 22Na outflux coefficient (Equation 8) K = Vici ' V0C0 ' (CRIHSAO. Co) DNa - C (38)(92.4) - (37)(82.3) - (2.3(82.3) 86 3.2 ul/min 142 4. Creatinine calculations a. mean ventricular concentration, E (Equation 5, Methods) = 1.79 + 0.37(2.21-1.79) = 1.95 ug/ul b. creatinine clearance, Ccr (Equation 5, Methods) = (38)(2.21) 55(37)(l.79) = 9.1 ul/min c. creatinine outflux coefficient, KDcr (Equation 8, Methods) = (39 (2.21) - (37) (1.79) - (2.3) Q.79) 1.95 = 7.0 ul/min d. creatinine flux rate, ficr (Equation 9, Methods) ncr = KDcr - c = Vici - Voc c ) (CRIHSA ° 0 0 _ (7.0)(1.95) = (38)(2.21) - (37)(1.79) - (2.3)(1.79) = 13.7 ug/min APPENDIX VIII EXTENDED FORTRAM PROGRAM FOR THE CALCULATION OF MASS BALANCE . ‘4‘!" .; ‘.I'I.Il'_ lu- APPENDIX VII I a moat .«c.«c.~«. ~h\n~\oa «uhao ~¢~4to6n> 2km oo:o coo m¢w40m. aqwqo.nu.m>.wtmhuma ..o«.«uu.an~a. .cm FZHWQ mm cc¢ Ch 00 « u.¥D .qupuua .o~.- pzuma m~ o» oo .~.am.oH . um new o» co . a.a“.on .un . ~¢ou .a.a«um .xo .«u .pczxou o~ .aa.auu..~.6..acwaomm.a¢u4o.au .u> .uzupuua .ou .nN mama a“ ca‘o oonmn Gama m u mzuaz .quw .xn: .x. mzo~»<4304co uoz¢ammHm: .ao u pawn >ooao~m>zu amzznn .x~m « .x .\ .«za .»¢raou Na .\.nH .wu¢azc uxowu .\ .«zd uhctflou Nu .m~.d u H .auuhe .wuduz .Nd h2aun .m~.« n H .anabu .wa brawn o u «mm: a O.ue¢¢2 « uucaz ad uxm mHIh mom Qanbwm hzuaa schm 000 a o» co . ~«m~ . a.a«uN .xo .«HN « . x~ . ouau 23m do ozu - «unaw: ace zo~ .p«:mou o .m~.dn~..n.»..pruuco.»uu3o .xcmz .oH .o pzuaa a“ o» ow . «.cu.oH .uH ooo o» oo . a.aw.aa .uH . ~¢m~ . ..o«u~ .xo .«HN.»¢zuou m .m~.au~..n.»..»:OH¢o.puwao .zamz .a» .m 9¢ux a aduo 02w no «wodw: Dcwa .a u tucker .a u aocdao .a u no¢zs .xm. . - umgcpo» uaazcm- - uzn~.x: . xcauhsozs ~ .xm .xoaupaozs.xm .onhowoaouxad .xm.maaxquo.x6 .uaazcmzm .xoa .\\.4»¢o waaxcmxau .xmo .bttaou mm on mm szaa a u uzuaz u mzHJz :n coo o» oo ~ u yucca m u mzHJ we a: o» oo . «.cu.x¢mz .uH raomnm 6 taupe» « taupe» H06a> 6 thpmwa n zoomam . \.~.~u .xsa..~.~u.xed.n.~.~u.x:a.\. mpaaze .xa~ . upcax: m .x- .uzoo:: .xuu. up¢uxc .xuu .mznpz: .xca .s 6 . zonhauomo< xaamxmu .xnu .wozqaawaoxo n .xod .xoguzuxo .xwa .xoauzuzo .xou .ooHuuazo .xmu ~ .\\. canawu wax» «on «hco ~z«hmzoox¢~ .xnm 6\\ « .~¢o« .xno . x . . 6 IN . no . \ chateau on ow wu J‘u eo:o Ono hwwImm tacoomm 145 moan 6«:6«:6Nd6 Ns\n~\oo \ 4400b 6 «oqdmo 6 o~u thaa 659a . 6 tcohohxdoqdau. u JJOUh . 6:626 466°» no bzuocwa znw .~.n«u .mrau :u « .~.o~u 6x66. 266mm 2H sqpopzaa . xua .6axuou oa~ A \ 6 .0 IN um . ~.n«a .mzau :m.~.o~u.x6 . ~.n66 .mzao xm.~.o~u.x6 bzuudm 6 human 6aa~ szma .aoa o a taopohxhuwdm . u bronxm Franco 9 huwam u huwao taoboh6 mud th¢a . m4<6op ozaauz~6 . . 6z~ .m .x~m .\ charred mow mow thOO «oN cam ch om m u xuamu e« u wzHJ ocN MdaFOp czamo hZHQQ 0 oxu no ozu 06 o» co .6 u acqcnm .6 u mucmsm .a u ~u 2hu oaJO ooo arc Oh on n I suamn hmuzmr 000 c o o tamocom mod and mm« emu mad cad wmu oNa mud 146 626 666 c» 66 6 6 62662 n 6266: mu 62666 6 xx .~666 .6666666taou 6~6 6 66.6 6 6.66.6. .6N6 62666 6N6 66666 . 666~ .666 .66 .66 .ow 6 o» oo 66m 6~6 o» 66 6 r.ou.y66c6 6 66 6m~.6 u 6.66.». .~6 62666 666.6 6 6.66.6. .w66az .~6 6z6¢a 6266 6 w 6 62662 66 66662 6 66667 666 66 66 6 xm266 .66 .6266z .66 6266 6 6266: u wz662 66o doxbzou mzHJ 024 made 6 66 66 6 6w¢ 2pm ca:o coo hwuzmm UOU tcduoun oa~ can mod ecu ms“ 956 BIBLIOGRAPHY 3". BIBLIOGRAPHY Allen, J. N. Extracellular Space in the central nervous system. Arch. Neurol. Psychiat., Chicago, 13:241-248, 1955. Ames, A., K. Higashi, and F. B. Nesbett. Effects of FCC acetazolamide and ouabain on volume and com- position of choroid-plexus fluid. J. Physiol. 181:516-524, 1965. 2’ Anderson, D. K. and R. L. Hazelwood. Chicken cerebrospinal fluid: normal composition and response to insulin administration. J. Physiol. 202:83-95, 1969. Baethmann, A., U. Steude, S. Horsch, and W. Brendel. The thiosulphate (355) Space in the CNS of rats after ventriculo-cisternal perfusion. Pflugers Arch. 316:51-63, 1970. Bering, E. A. and O. Sato. Hydrocephalus: changes in forma- tion and absorption of cerebrospinal fluid within the cerebral ventricles. J. Neurosurg. 22:1050-1063, 1963. Bierer, D. W. Active transport of organic anions from the brain ventricles of the dog (Master's Thesis), Michigan State University, East Lansing, Michigan. Bito, L. 2. Blood brain barrier: evidence for active cation transport between blood and the extracellular fluid of brain. Science 165:81-83, 1969. Bourke, R. S., E. S. Greenberg, and D. B. Tower. Variation of cerebral cortex fluid spaces in vivo as a func- tion of Species brain Size. Am.'31 Physiol. 208: 682-692, 1965. Bradbury, M. W. B. and H. Davson. The transport of urea, creatinine and certain monosaccharides between blood and fluid perfusing the cerebral ventricular system Of rabbits. J. Physiol. 110:195-211, 1964. 147 148 Bradbury, M. W. B. and H. Davson. The transport of otas- Sium between blood, cerebrOSpinal fluid and rain. J. Physiol. 181:151-174, 1965. Bradbury, M. W. B. and B. Stulcova. Efflux mechanism con- tributing to the stability of the potassium concen- tration in cerebrospinal fluid. J. Physiol. 208: 415-430, 1970. _—_ Cicardo, V. H. The mechanism of arterial hypertension after potassium injection into the nerve centers. Arch. Int. Pharmacodyn. 80:199-208, 1949. Cicardo, V. H. Arterial hypertension obtained by cisternal injections of sodium citrate. Arch. Int. Pharma- codyn. §§;l74-181, 1950. Cohen, D. H., A. M. Schnall, R. L. Macdonald, and L. H. Pitts. Medullary cells of origin of vaga1.cardioinhibitory fibers in the pigeon. I. Anatomical studies of peripheral vagus nerve and the dorsal motor nucleus. J. Comp. Neur. 140:299-320, 1970. Cohen, D. H. and A. M. Schnall. Medullary cells of origin of the vagal cardioinhibitory fibers in the pigeon. II. Electrical stimulation of the dorsal motor nucleus. J. Comp. Neur. 140:321-342, 1970. COOper, E. S., S. Holm and S. Bellet. The cardiovascular effects following intracisternal potassium chloride infusion. Am. J. Med. Sci. 236:705-719, 1958. Cserr, H. Potassium exchange between cerebrOSpinal fluid, plasma, and brain. Am. J. Physiol. 209:1219-1226, 1965. Cutler, R. W. P., L. Page, J. Galicich, and G. V. Watters. Formation and absorption of cerebrospinal fluid in man. Brain 91:707-720, 1968. Cutler, R. W. P., R. J. Robinson, and A. V. Lorenzo. Cere- brospinal fluid transport Of sulfate in the cat. Am. J. Physiol. 214:448-454, 1968. Dandy, W. E. Experimental hydrocephalus. Ann. Surg. 10: 129-142, 1919. Davson, H. Physiology of the Cerebrospinal Fluid. J. and A. Churchill Ltd., London, 1967. 149 Davson, H., C. R. Kleenman and E. Levin. Quantitative studies of the passage of different substances out of the cerebrospinal fluid. J. Physiol. 161: 126-142, 1962. '——' Davson, H. and M. B. Segal. Effect of cerebrospinal fluid on volume of distribution of extracellular markers. Brain 92:131-136, 1969. Davson, H. and E. Spaziani. The blood-brain barrier and the extracellular Space of brain. J. Physiol. 149: 135-143, 1959. Devos, J. Central influences on heart rate. Arch. Int. Physiol. 52:385-387, 1951. Dowben, R. M. Cell Biology. Harper and Row, New York, 1971. Downmann, C. B. B. and C. C. Mackenzie. Intracisternal injection of potassium phosphate. Lancet 245:471- 473, 1943. Elliot, K. A. C. and H. Jasper. Measurement of experimen- tally induced brain swelling and Shrinkage. Am. J. Physiol. 157:122-129, 1949. Emerson, T. E. and W. H. Massion. Direct and indirect vascular effects of cycloprOpane and halothane in the dog forelimb. J. Appl. Physiol. 22:217-222, 1967. "' Feldberg, W. and S. L. Sherwood. Effects of calcium and potassium injected into the cerebral ventricles of the cat. J. Physiol. 139:408-416, 1957. Fishman, R. A. Carrier transport of glucose between blood and cerebrospinal fluid. Am. J. Physiol. 206:836- 844, 1964. Flexner, L. B. A note on the rate of circulation of cere- brospinal fluid. Am. J. Physiol. 106:201-202, 1933. Flexner, L. B. and H. Winters. The rate of formation of cerebrospinal fluid in etherized cats. Am. J. Physiol. 101:697-710, 1932. Frazier, C. H. and M. M. Peet. Factors of influence in the origin and circulation of the cerebrospinal fluid. Am. J. Physiol. 35:268-282, 1914. Giese, A. C. Cell Physiology, 3rd edition. W. B. Saunders Co., Philadelphia, 1968. 150 Goodman, L. S. and A. Gilman, eds. The Pharmacological Basis of Therapeutics. 3rd edition. Macmillan Co., New York, 1965. Graziani, L., A. Escriva, and R. Katzman. Exchange of calcium between blood, brain, and cerebrospinal fluid. Am. J. Physiol. 208:1058-1064, 1965. Graziani, L. J., R. K. Kaplan, A. Escriva, and R. Katzman. Calcium flux into CSF during ventricular and ventri- culocisternal perfusion. Am. J. Physiol. 213:629- 636, 1967. Greenberg, D. M., R. B. Aired, M. D. D. Boelter, W. W. Campbell, W. E. Cohn, and M. M. Murayama. A study with radioactive isotopes of the permeability of the blood-cerebrospinal fluid barrier to ions. Am. J. Physiol. 140:47-64, 1943. Hammerstad, J. P., A. V. Lorenzo, and R. W. P. Cutler. Iodide tranSport from the Spinal subarachnoid fluid in the cat. Am. J. Physiol. 216:353-358, 1969. Hansen-Pruss, O. C. Meninges of birds with consideration of the Sinus rhomboidalis. J. Comp. Neurol. 36: 193-216, 1923. ‘— Heisey, S. R. Cerebrospinal and extracellular fluid Spaces in turtle brain. Am. J. Physiol. 219:1564-1567, 1970. Heisey, S. R., D. Held, and J. R. Pappenheimer. Bulk flow and diffusion in the cerebrospinal fluid system of the goat. Am. J. Physiol. 203:775-781, 1962. Heisey, S. R. and D. K. Michael. Cerebrospinal fluid forma- tion and bulk absorption in the freshwater turtle. Exp. Neurol. 31:258-262, 1971. Held, D., V. Fencl, and J. R. Pappenheimer. Electrical potential of cerebrospinal fluid. J. Neurophysiol. 21:942-959, 1964. Heymans, C. and E. Neil. Reflexogenic Areas of the Cardio- vascular System. J. and A. Churchill Ltd., London, 1958. Hochwald, G. M. and M. Wallenstein. Exchange of albumin between blood, cerebrospinal fluid and brain in the cat. Am. J. Physiol. 212:1199-1204, 1967. 151 Hooker, D. R. The perfusion of the mammalian medulla: the effect of calcium and of potassium on the respiratory and cardiac centers. Am. J. Physiol. 38:200-208, 1915. ‘—— Horsten, G. P. M. and P. J. KlOpper. Effects of changes in compostion of the cerebrospinal fluid on the elec- troencephalogram. Arch. Int. Physiol. 60:491-505, 1952. Huggins, C. B. and A. B. Hastings. Effect of calcium and citrate injections into cerebrospinal fluid. Proc. Soc. Exptl. Biol. Med. 30:459-460, 1933. Katzman, R. Effect of electrolyte disturbance on the central nervous system. Ann. Rev. Med. 11:197-212, 1966. Katzman, R., L. Graziani, R. Kaplan, and A. Escriva. Exchange of cerebrOSpinal fluid potassium with blood and brain. Arch. Neurol. 13:513-524, 1965. Katzman, R. and F. Hussey. A Simple constant-infusion manometric test for measurement of CSF absorption. Neurology 29:534-544, 1970. Katzman, R. and P. H. Leiderman. Brain potassium exchange in normal adult and immature rats. Am. J. Physiol. 175:263-270, 1953. Last, R. J. and D. H. Tompsett. Casts of the cerebral ven- tricles. Brit. J. Surg. 40:525-543, 1953. Leiderman, P. H. and R. Katzman. Effect of adrenalectomy, desoxycorticosterone, and cortisone on brain potas- Sium exchange. Am. J. Physiol. 175:271-275, 1953. Leusen, I. The influence of calcium, potassium, and mag- nesium ions in cerebrospinal fluid on vasomotor system. J. Physiol. 110:319-329, 1950. Lorenzo, A. V., J. P. Hammerstad, and R. W. P. Cutler. Cerebrospinal fluid formation and absorption and transport of iodide and sulfate from the spinal subarachnoid space. J. Neurol. Sci. 12:247-258, 1970. Marshall, A. J. ed. Biology and Comparative Physiology of Birds. Volume I. ACademic Press, New York, 1960. Masserman, J. H. Cerebrospinal hydrodynamics. Arch. Neurol. Psychiat. (Chicago). 32:523-553, 1934. 152 Maynard, E. A., R. L. Schultz, and D. C. Pease. Electron micrOSCOpy of the vascular bed of the rat cerebral cortex. Am. J. Anat. 100:409-433, 1957. McGinnis, C. H. and R. K. Ringer. Carotid sinus reflex in the chicken. Poultry Sci. 453402-404, 1966. Merlis, J. K. The effect of changes in the calcium content of the cerebrospinal fluid on Spinal reflex activity in the dog. Am. J. Physiol. 131:67-72, 1940. Milhorat, T. H. Choroid plexus and cerebrospinal fluid production. Science 166:1514-1516, 1969. Millen, J. W. and D. H. M. Woollam. The Anatomy of the CerebrOSpinal Fluid. Oxford University Press, London, 1962. Mullins, F. J., A. B. Hastings, and W. H. Lees. Neuro- muscular responses to variations in calcium and potassium concentrations in the cerebrospinal fluid. Am. J. Physiol. 1215719-727, 1938. Murry, J. E. and R. W. P. Cutler. Transport of glycine from the cerebrOSpinal fluid. Arch. Neurol. 23:23-31, 1970. - Natelson, S. Microtechniques of Clinical Chemistry. 2nd edition. C. C. Thomas, Springfield, 1961. Oppelt, W. W., E. S. Owens, and D. P. Rall. Calcium exchange between blood and cerebrospinal fluid. Life Sciences 8:599-605, 1963. Oppelt, W. W., C. S. Patlak, C. G. Zubrod, and D. P. Rall. Ventricular fluid production rates and turnover in elasmobranchii. Comp. Biochem. Physiol. 12:171-177, 1964. '_' Pappenheimer, J. R., V. Fencl, S. R. Heisey, and D. Held. Role of cerebral fluids in control of respiration as studied in unanesthetized goats. Am. J. Physiol. 208:436-450, 1965. Pappenheimer, J. R., S. R. Heisey, and E. F. Jordan. Active transport of Diodrast and phenolsulfonphthalein from cerebrospinal fluid to blood. Am. J. Physiol. 220: 1-10, 1961. 153 Pappenheimer, J. R., S. R. Heisey, E. F. Jordan, and J. DeC. Pollay, Pollay, Pollay, Rall, D. Reed, D. Rennick, Resnik, Rodbard, Rodbard, Downer. Perfusion of the cerebral ventricular system in unanesthetized goats. Am. J. Physiol. 203:763-774, 1962. M. Cerebrospinal fluid transport and the thiocyanate Space of the brain. Am. J. P ysiol. 210:275-279, 1966. M. and F. Curl. Secretion of cerebrospinal fluid by the ventricular ependyma of the rabbit. Am. J. Physiol. 213:1031-1038, 1967. M. and R. J. Kaplan. Effect of the CSF sink on thiocyanate concentration gradient in brain. Am. J. Physiol. 219:802-808, 1970. P., W. W. Oppelt, and C. S. Patlak. Extracellular Space of brain as determined by diffusion of inulin from the ventricular system. Life Sciences 1:43-48, 1962. J. and D. M. Woodbury. Kinetics of movement of iodide, sucrose, inulin, and radio-iodinated serum albumin in the central nervous system and cerebro- spinal fluid of the rat. J. Physiol. 169:816-850, 1963. B. R. Transport mechanisms for renal tubular excretion of creatinine in the chicken. Am. J. Physiol. 212:1131-1134, 1967. H., Jr., M. F. Mason, R. T. Terry, C. Pilcher, and T. R. Harrison. The effect of injecting certain electrolytes into the cisterna magna on the blood pressure. Am. J. Med. Sci. 191:835-850, 1936. 8. Evidence for a temperature-sensitive center in a poikilotherm, the turtle. Fed. Proc. 6:191, 1947. S. and M. Tolpin. A relationship between the body temperature and the blood pressure in the chicken. Am. J. Physiol. 151:509-515, 1947. de Rougemont, J., A. Ames III, F. B. Nesbitt, and H. F. Hofmann. Fluid formed by choroid plexus. A tech- nique for its collection and a comparison of its electrolyte composition with serum and cisternal fluids. J. NeurOphysiol. 23:485-495, 1960. 154 Rubin, M. A., H. E. Hoff, A. W. Winkler, and P. K. Smith. Intravenous potassium, calcium and magnesium and the cortical electrogram of the cat. J. Neuro- physiol. 6:23-28, 1943. Rubin, R. G., E. 8. Henderson, A. K. Ommaya, M. D. Walker, and D. P. Rall. J. Neurosurg. 155430-438, 1966. Sahar, A., G. M. Hochwald, and J. Ransohoff. Experimental hydrocephalus: cerebrospinal fluid formation and ventricular Size as a function of intraventricular pressure. J. Neurol. Sci. 11:81-91, 1970. Sahar, A., G. M. Hockwald, and J. Ransohoff. Cerebrospinal fluid turnover in experimental hydrocephalic dogs. Neurology 11:218-224, 1971. Shannon, J. A. The excretion of exogenous creatinine by the chicken. J. Cell Comp. Physiol. 11:123-134, 1938. "' Smith, H. W. Principles of Renal Physiology. Oxford University Press, New York, 1956. Smolik, B. Effect of intracisternal injection of potassium phosphate in hemorrhagic hypotension and Shock in the dog. Proc. Soc. Exptl. Biol. Med. 11:70, 1943. Snodgrass, R. S., R. W. P. Cutler, E. S. Kang, and A. V. Lorenzo. Transport of neutral amino acids from feline cerebrospinal fluid. Am. J. Physiol. 111: 974-980, 1969. Stern, L. Direct chemical action upon nerve centers in biology and medicine. Nature 156:7-9, 1945. Sturkie, P. D., ed. Avian Physiology. 2nd edition. Cornell University Press, New York, 1965. Sweet, W. H. and H. B. Locksley. Formation, flow and reab- sorption of cerebrospinal fluid in man. Proc. Soc. Exptl. Biol. Med. 81:397-402, 1953. Sykes, A. H. Excretion of inulin, creatinine and ferro- cyanide by the fowl. Res. Vet. Sci. 1:315-320, 1960. Truex, R. C. and M. B. Carpenter. Human Neuroanatomy. 6th edition. Williams and Wilkins Co., Baltimore, 1969. 155 Tschirgi, R. D. Handbook of Physiology. Neurophysiology III. 1865-1890, Williams and Wilkens Co., Baltimore, 1960. Van Harreveld, A. Asphyxial changes in the cerebral cortex. J. Cell. Comp. Physiol 11:101-110, 1961. Van Harreveld, A., N. Ahmed, and D. J. Tanner. Sulfate concentrations in cerebrospinal fluid and serum of rabbits and cats. Am. J. Physiol. 210:777-780, 1966. Van Harreveld, A., J. Crowell, and S. K. Malhotra. A study Of extracellular Space in central nervous tissue by freeze substitution. J. Cell. Biol. 25:117-137, 1965. ‘- Walker, S. M., E. A. Smolik, and A. S. Gilson. The effects of intracisternal injection of potassium phosphate on the rate and rhythm of the heart and on the blood pressure and on the respiration of the dog. Am. J. Physiol. 145:223-238, 1945. Wallace, G. B. and B. B. Brodie. On the source of cerebro- Spinal fluid. The distribution Of bromide and iodide throughout the central nervous system. J. Pharmacol. Exp. Therap. 19:418-427, 1940. Weed, L. H. The pathways of escape from the subarachnoid Space with particular reference to the subarachnoid villi. J. Med. Res. 11:51-91, 1914. Welch, K. Secretion of cerebrospinal fluid by choroid plexus of the rabbit. Am. J. Physiol. 205:617-624, 1963. Welch, K. and V. Friedman. The cerebrospinal fluid valves. Brain 81:454-469, 1960. Wellman, H. N., H. P. Lewis, R. Carroll, and R. Rameriz. Visualization and functigg studies of the cerebro- spinal fluid Space with mTc sulfide colloid. Neurology 18:1113-1119, 1968. Whittow, G. C., P. D. Sturkie, and G. Stein Jr. Cardio- vascular changes in restrained chickens. Poultry Sci. 41:1452-1459, 1965. 156 Winterstein, H. The action of substances introduced into the cerebrOSpinal fluid and the problem of intra- cranial chemoreceptors. Pharmacol. Rev. 13:71-107, 1961. “— Woodward, D. L., D. J. Reed, and D. M. Woodbury. Extra- cellular Space Of rat cerebral cortex. Am. J. Physiol. 212:367-370, 1967. IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII WWI mmmuquwllmlwwfllmlu 293 03