UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Dilution acidosis : the effects of hyperosmolality on acid-base balance and ventilation Kasserra, Claudia E. 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1993_spring_phd_kasserra_claudia.pdf [ 3.72MB ]
JSON: 831-1.0086340.json
JSON-LD: 831-1.0086340-ld.json
RDF/XML (Pretty): 831-1.0086340-rdf.xml
RDF/JSON: 831-1.0086340-rdf.json
Turtle: 831-1.0086340-turtle.txt
N-Triples: 831-1.0086340-rdf-ntriples.txt
Original Record: 831-1.0086340-source.json
Full Text

Full Text

DILUTION ACIDOSIS: THE EFFECTS OF HYPEROSMOLALITY ON ACID-BASEBALANCE AND VENTILATIONbyCLAUDIA EVE KASSERRAB.Sc., The University of Guelph, 1984A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FORTHE DEGREE OF DOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Zoology)We accept this thesisas conforming to the required standardTHE UNIVERSITY OF BRITISH COLUMBIADecember 1992© Claudia Eve KasserraIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of  Zool ogyThe University of British ColumbiaVancouver, CanadaDate^March 25, 1993DE-6 (2/88)iiABSTRACTThe effects of acute osmotic changes on acid-base balance and respiratory controlwere studied in the Pekin duck, Anas platyrhynchos. Acute hyperosmolality followingintravascular injection of essentially non-penetrating solutes such as NaCl or sucrosecaused an increase in extracellular fluid volume, a prolonged extracellular acidosis (so-called dilution acidosis), and a relative increase in extracellular Cl- concentration. Incontrast, hyposmolality did not cause complementary changes in these variables. Studiesin acute hyperosmotic stress were therefore undertaken to investigate both the nature ofthe acid-base disturbance and the implications for ventilatory control. 31P nuclearmagnetic resonance spectroscopy (31P NMR) on duck pectoral muscle showed that thedilution acidosis caused by acute hyperosmolality was accompanied by an intracellularcontraction alkalosis, implying the uncoupling of intra- and extracellular pH. Theincrease in extracellular [CF] and the pH changes suggested a primary role for CF/HCO3 -exchange during this perturbation. However, both the anion-exchange blocker DIDS andthe Na+/H+ exchange blocker amiloride reversed the intracellular pH change fromalkalosis to acidosis, although they did not affect the extracellular acidosis caused by acutehyperosmolality. These results indicated that hyperosmolality altered both Cl/HCO3 - andNa+/H+ exchange and also suggested the involvement of one or more additional exchangemechanisms.Despite the pronounced extracellular acidosis during hyperosmolality, there was nocompensatory stimulation of ventilation. An extracellular pH decrease of similarmagnitude and time course when caused by lactic acid infusion stimulated a 100% increasein ventilation, and decreased both extracellular and intracellular pH in contrast to theintracellular alkalosis during hyperosmolality. Although intracellular pH was not directlymeasured in chemoreceptor tissue, it is reasonable to assume that such well-perfused tissuewould be exposed to any osmotic stress. Reversal of the intracellular alkalosis duringiiihyperosmolality by DIDS and amiloride, so that both extra- and intracellular pH wereacidotic, resulted in a significant increase in ventilation. These data are unique, since it isthe first piece of clear, although indirect, evidence that intracellular pH plays a role ininitiating ventilatory changes to acid-base disturbances at the peripheral chemoreceptors.Since brain intracellular pH as measured by 31P NMR during systemic hyperosmolalityshowed only a consistent trend towards an alkalosis, then central chemoreception, if basedon intracellular pH, is unlikely to be affected by hyperosmolality. Acute hyperosmolalityalso increased the ventilatory threshold to acute hypercapnia, reaffirming a depression ofperipheral chemoreception. However, the ventilatory threshold to hypoxia was decreasedand sensitivity to K+ was increased, indicating that the chemoreceptive mechanisms forCO2 and 02 are different, and that both intra- and extracellular pH are crucial toventilatory control.ivTABLE OF CONTENTSABSTRACT^ iiTABLE OF CONTENTS^ ivLIST OF TABLES viiLIST OF FIGURES^ viiiACKNOWLEDGEMENTSINTRODUCTION^ 1MATERIALS AND METHODS^ 7Surgical Methods 7Physiological Responses to Acute Changes in Osmolality^7Hyperosmolality^ 7Experimental Protocol^ 7Hyposmolality^ 10Experimental Protocol^ 10Hyper- and Hyposmotic Calculations and Data Analysis^113 IP Nuclear Magnetic Resonance Spectroscopy^ 11Experimental Protocol^ 113 IP NMR Spectra of Pectoral Muscle^ 1231P NMR Spectra of Brain^ 1331P NMR Calculations and Data Analysis^ 1631P NMR of Pectoral Muscle During Ion-Exchange Blockade^19Experimental Protocol^ 19Calculations and Data Analysis^ 20Responses to Respiratory Stimuli 20Experimental Protocol^ 20Calculations and Data Analysis^ 22VRESULTS^ 23Physiological Responses to Acute Changes in Osmolality^23Hyperosmolality versus Lactate^ 23Hyposmolality^ 3231P Nuclear Magnetic Resonance Spectroscopy^ 32Pectoral Muscle^ 32Brain^ 35Ion-Exchange Blockade^ 37DIDS^ 37Amiloride 37Total CO2^ 44Responses to Respiratory Stimuli^ 45Hypercapnia^ 45Hypoxia 49Potassium^ 54DISCUSSION^ 57Physiological Responses to Acute Changes in Osmolality^57Analysis of Acid-Base Status^ 57Constancy of Plasma [K+] 58Respiratory Compensation^ 59Summary^ 6231P NMR of Muscle and Brain^ 62Peripheral Chemoreception 63Central Chemoreception^ 64Metabolism^ 66Summary 6631 13 NMR of Muscle During Ion-Exchange Blockade^67viDosage^ 67Control of Ventilation^ 68Ion-Exchange in Single Cells 71Total CO2^ 72Summary 72Responses to Respiratory Stimuli^ 74Hypercapnia^ 74Hypoxia and 1(± 76Metabolism^ 78Summary 78CONCLUSIONS^ 79REFERENCES 80V i iLIST OF TABLESTable I. Ionic, pHa and blood gas changes 30 min after the start of hyperosmoticinfusion.^ 26Table II. Decreases in electrolytes during acute hyperosmolality alone andhyperosmolality after various treatments.^ 73ViiiLIST OF FIGURESFigure 1. Schematic diagram of the experimental setup for all studies involvingplethysmography.^ 8Figure. 2. Four successive spectra of duck pectoral muscle.^15Figure. 3. Representative spectra of duck brain with skull and overlying skin intact. 18Figure. 4. Changes observed in osmotic pressure, plasma [Nal, [C11, and [K+]during control (isotonic saline), 4 M NaCl and 2.4 M sucrose infusions.^25Figure. 5. The changes observed in pHa, VE, PaCO2 and Pa02 during control(isotonic saline), 4M NaC1, 2.4M sucrose and lactate infusions.^29Figure. 6. The changes in pHa, 7E, and PaCO2 measured in a single animal duringinfusion of 2.4 M sucrose (7.5 mmol•kg -1 ).^ 31Figure. 7. Changes in pH and blood gases during hypertonic and lactic acidinfusions.^ 34Figure. 8. Changes in arterial pH and brain intracellular pH during hypertonicsucrose infusion.^ 36Figure. 9. The effect of DIDS treatment on the pH and blood gas response tohyperosmolality.^ 39ixFigure 10. Changes in ventilation in response to hyperosmolality during ionexchange blockade.^ 40Figure 11. Changes in plasma [Nat] and [C1-] during hyperosmolality and ionexchange blockade.^ 41Figure 12. The effect of amiloride treatment on the pH and blood gas response tohyperosmolality.^ 43Figure 13. pHa, blood gas and ventilatory changes in response to hypercapnia andsucrose.^ 47Figure 14. Changes in VO2 associated with hypercapnia and hypoxia before andduring hyperosmotic stress.^ 48Figure 15. CO2 and 02 ventilatory response curves.^ 50Figure 16. pHa, blood gas and ventilatory changes in response to hypoxia andsucrose.^ 53Figure 17. The respiratory response to bolus K+ infusion.^56Figure 18. The relationship between ventilation (VE) and intra- and extracellular pHduring hyperosmolality.^ 69ACKNOWLEDGMENTSI would like first of all to thank my supervisor, David R. Jones, who, more thananybody else, toughened me up, and who offered me support and encouragement in hisown unique way. Many thanks to my committee members, John Gosline, MaryanneHughes, John Ledsome, and John Steeves for their careful perusal of this work and forletting me just get on with it during the experiments. The fact that the time I spent herewas so much fun was due to the warped senses of humour and unwavering support of allthe friends I made here, especially Tim West, Mike Hedrick, David McClellan, JimStaples, Sohail Hasan, Agnes Lacombe, and Greg Funk. I am indebted to each of them.I would particularly like to say thank you to Chris Moyes, not only for his help and all thelate night discussions about the meaning of science (the answer is 42), but for his love andtremendous moral support over the years. I am also grateful to my parents, who nevershared my doubts and always came through with the help I needed. Finally, I thank myhusband-to-be, Christopher Courtin, for, well, just everything.I would like to acknowledge NSERC and UBC for financial support over theduration of these studies.1INTRODUCTIONChanges in extracellular osmolality are often accompanied by changes inextracellular and intracellular water volumes, electrolyte concentrations, and acid-baseparameters. Acute hyperosmolality due to essentially non-penetrating solutes such assucrose or NaCl leads to a prolonged extracellular acidosis, which has been termeddilution acidosis (Shires and Holman, 1948; Sotos et al., 1962; Makoff et al., 1970).This is an unusual effect because it occurs without the addition or depletion of either anacid or a base. Dilution acidosis has been examined cursorily in mammals, and a similarplasma acidosis has been demonstrated in teleosts and amphibians in response to ahypertonic environment (Wilkes and McMahon, 1986; Walker et al., 1990). Althoughlarge, acute changes in osmolality are not normally experienced by many vertebratesexcept humans and certain fish, acute osmotic increases in humans are a seriouspathophysiological condition experienced during illness such as diabetes and in treatmentsinvolving hemodialysis or altered electrolyte retention. Despite the clinical implications,little work has been done on this phenomenon. Much more work has been done on theantithesis of dilution acidosis, namely contraction (or Cl - deficiency ) alkalosis (Galla andLuke, 1988; Wesson, 1990; Galla et al., 1984). The conventional explanation of dilutionacidosis is based upon the dilution of plasma by the movement of intracellular water to theextracellular space in response to the osmotic gradient created by hypertonic infusion.Due to the continuous tissue production of CO2 and its ready diffusion across cellmembranes (therefore PaCO2 would not change), the dilution would only measurablyreduce bicarbonate concentration, resulting in a drop in extracellular pH (pHe).Bicarbonate generation and equilibration may be inhibited during extracellularhyperosmolality because of perturbation of intermediary cell metabolism (Chang et al.,1975; Makoff et al., 1970), although this is controversial. It has also been suggested thatbicarbonate transfer out of cells is instead augmented (Winters et al., 1964).2The development of the extracellular acidosis is apparently accompanied byconcomitant development of an intracellular alkalosis, which implies the uncoupling ofintra- and extracellular pH. There have been numerous demonstrations of various tissuesand single cells in vitro becoming alkalotic upon exposure to a hypertonic medium,including skeletal muscle (Adler et al, 1975; Abercrombie and Roos, 1983), cardiacmuscle (Whalley et al., 1991), glial cells (Jean et al., 1986), and osteoblasts (Green etal., 1988), but there is a paucity of information obtained in vivo. Makoff et al. (1970)reported a significant increase in intracellular pH (pHi) of erythrocytes in anesthetizeddogs (pump ventilated, nephrectomized, splenectomized) 2 hr after infusion withhypertonic NaC1 or mannitol, but it is not known whether this response occurs in anyother tissue. There are likely to be significant differences in the response to osmotic stressbetween cells which are excellent cell volume regulators, such as erythrocytes, and thosethat are not, such as skeletal muscle.Dilution of the extracellular space by infusion of isotonic or hypotonic fluidsgenerally does not result in a measurable acidemia (see Garella et al., 1975), althoughrapid infusion of massive volumes of isotonic fluid can transiently increase the hydrogenion concentration ([H+]; Shires and Holman, 1948; Rosenbaum et al., 1969). Infusion offreely diffusible solutes, such as urea, which do not cause a shift of water from theintracellular compartment or affect pHi (Adler et al., 1975), also have no effect on pHe(Winters et al., 1964). This suggests that the crenation, or the increase in intracellularosmolality or possibly pH, during exposure to hypertonic solutions somehow altersmembrane transport. While most of the studies in single cells implicate Na+/H+ exchangein the response to a hypertonic medium (Green and Muallem, 1989; Whalley et al.,1990), in vivo studies suggest a predominant role for anion exchange (Makoff et al.,1970; Sotos et al., 1962). These in vivo studies show a relative increase in extracellular[Cr], implying an outward flux of CV. If the CV were exchanged for HCO3 - , this couldresult in an extracellular acidosis and intracellular alkalosis. Use of selective ion-exchange3blockers in vivo would help to resolve which ion-exchange perturbations occur in responseto hyperosmolality.Normally, extracellular acid-base disturbances elicit compensatory changes inrespiration. The degree of respiratory compensation has never been measured duringdilution acidosis, but the literature shows anomalous changes in blood gases. In studieswith rabbits and dogs, which were allowed to breathe normally during dilution acidosis,PaCO2 either decreased only very slightly or actually increased, even after a fall in arterialpH (pHa) of 0.3 pH units, suggesting a lack of ventilatory compensation (Sotos et al.,1962; Asano et al., 1966). The same decrease in pHa caused by infusion of acidsapproximately doubled ventilation (TE) in rabbits (Maskrey and Trenchard, 1989; Nattie,1983) and decreased PaCO2 8 Torr (Nattie, 1983). It has long been hypothesized thatintracellular pH changes may be involved in chemoreception (in addition to extracellularchanges) (Hanson et al., 1981; Adler et al., 1990; Lassen, 1990), but there is noconclusive evidence regarding this hypothesis to date. Suppression of the normalchemoreceptor response to acidosis could be explained if an intracellular contractionalkalosis developed concomitantly, and the pH sensitive component of the chemoreceptorresponse was at least partially dependent upon intracellular conditions. The testing of sucha hypothesis would involve ascertaining the response of both peripheral and central tissuesto hyperosmotic stress. There is currently very little information on the response of thebrain to acute, systemic, hyperosmotic stress. The brain is a tissue which cell volumeregulates to some degree (see Macknight et al., 1992) and is partially protected by theblood-brain barrier.Chemoreceptors in birds are located both peripherally and centrally, but knowledgeof them is not nearly as extensive as in mammals. Peripheral receptors are locatedprimarily in the carotid bodies (Jones and Purves, 1970a) and the intrapulmonarychemoreceptors (Fedde et al., 1974), the afferents of which travel in the carotid sinus andvagus nerves, respectively, while central receptors exist somewhere caudal to the4mesencephalon (see Scheid and Piiper, 1986). Carotid chemoreceptors are sensitive to02, and all three receptor groups are sensitive to pH and CO 2 (see Scheid and Piiper,1986 for review). Transduction in the chemoreceptors has not yet been conclusivelydetermined. The majority of the work has been done on the carotid bodies, on which thisbrief review will be based (for a recent, more complete review, see Gonzalez et al.,1992). The most accepted proposed model for transduction of acidic stimuli (Rocher etal., 1991) suggests that an increase in [Hli stimulates operation of Na+-coupled H+-extruding systems that increase [Nali. The increase in intracellular Na+ drives the entryof Ca2 + by the Na±/Ca2+ exchanger, which activates exocytosis and release of dopamine.Dopamine is the neurotransmitter that then activates the apposed afferent nerve terminals(Rigual et al., 1991). However, the kinetics of this model have not been worked out, norhas the relationship between intra- and extracellular pH been established. Thechemotransduction of hypoxia is less clear and has been proposed in two models. Theplasma membrane model suggests that low P02 inhibits K+ channels, producing an initialdepolarization that activates voltage-dependent Ca 2+ channels and allows entry of Ca 2 +.Simultaneous activation of Na+ channels allows faster recruitment of Ca2 + channels,thereby increasing the entry of Ca2 + and subsequent release of neurotransmitters (Lopez-Lopez et al., 1989; Urena et al., 1989). The nature of the 02 sensor in this model isunknown, and there is considerable controversy over the relationship between carotid bodyP02 and the sensitivity of the K+ channels. The second model that has been proposed isthe metabolic model, in which hypoxia results in a decrease in carotid bodyphosphorylation potential, which somehow is a signal to decrease neurotransmitter release(originally proposed by Anichkov and Belenkii, 1963). A recent version of this modelsuggests that the actual signal for chemoreception in hypoxia is the slowing of the electrontransfer in the respiratory chain of chemoreceptor cells, resulting in a decrease in themitochondrial H± electrochemical gradient. This causes release of Ca 2+ from themitochondria, a rise in [Ca2 +]i and release of the neurotransmitter dopamine (Biscoe and5Duchen, 1990a). However, there are numerous criticisms of the metabolic model.Cytochrome oxidase has a very high affinity for 02 while the carotid bodies have a veryhigh tissue P02 . This would require the presence of a low-affinity cytochrome oxidase,the existence of which has been questioned (Acker and Eyzaguirre, 1987). Secondly, it isdifficult to reconcile a decrease in the phosphate potential being the signal for hypoxicchemoreception since that is the time of maximal function of the carotid body (Acker etal., 1989). Finally, the ability of mitochondria to sufficiently elevate cytosolic Ca 2+ isquestionable (McCormack et al., 1990).Intracellular pH is traditionally a difficult measurement to make repetitively withboth accuracy and precision. The use of classical techniques, such as freeze fracture orthe distribution of weak acids like dimethadione, all have a number of seriousshortcomings. All of these methods are highly invasive, do not easily lend themselves torepeated measurements, and may contain large error since the final calculation of pHidepends upon the measurement of several other variables, each associated with someerror. Intracellular pH, however, can presently be measured in many tissues, includingmuscle and brain, by 31-phosphorus nuclear magnetic resonance spectroscopy (31P NMR),which has several advantages as an analytical tool. It is non-invasive, allows repeatedmeasurements over time in a single animal, and follows cellular energy homeostasis bymonitoring changes in high energy phosphate metabolites and consequently, pHi. 31PNMR has a resolution of pHi measurement at least equal to conventional biochemicaltechniques.The experiments in this thesis were done with the domestic Pekin duck, Anasplatyrhynchos, because of its tolerance to large osmotic changes. The purpose of thisthesis was to identify the ionic changes leading to acid-base disturbance during acuteplasma osmolality changes, and to relate respiratory changes to specific compartmental pHchanges. It was hypothesized that acute hyperosmolality would result in both anextracellular dilution acidosis and an intracellular contraction alkalosis, and that the6intracellular alkalosis would suppress the increase in ventilation normally associated withacidosis. Furthermore, the pH perturbations were hypothesized to be primarily a result ofaltered C1-/HCO3 - exchange. The study first characterized the ionic, cardiovascular,respiratory and acid-base response to acute hypo- and hyperosmolality, and compared theresponse to a more conventional acidosis (lactacidosis). The investigation continued bydetermining the intracellular response of both brain and systemic muscle to theextracellular acidosis generated by either acute hyperosmolality or lactic acid infusion inthe conscious animal, thus establishing a relationship between different compartmentalchanges in pH to changes in ventilation. Since it is not feasible to monitor pHi within acarotid body in vivo by NMR (the primary location of the peripheral chemoreceptors), wemeasured pHi in skeletal muscle. The assumption that the effect of hypertonicity issimilar in the pectoral muscle and carotid body tissue is reasonable, since both tissues arewell perfused (De Castro and Rubio, 1968) and would therefore be similarly exposed toany changes in tonicity, and these tissues do not volume regulate extensively under acutehyperosmotic conditions (see Macknight et al., 1992). Specific ion-exchange blockadewas used during hyperosmotic stress in an attempt to prevent the pH perturbations, and todefine the ion-exchange mechanisms involved. We used 4,4'-diisothiocyanostilbene-2,2 1 -disulfonic acid (DIDS) to inhibit carrier-mediated anion exchange, including C1 -/HCO3 -exchange, and amiloride to inhibit the Na+/H+ antiporter. Since a suppression of thenormal ventilatory response to dilution acidosis would imply some degree of depression ofchemoreceptor discharge at least in the periphery, this was tested by examining theventilatory response to several known respiratory stimuli (hypercapnia, hypoxia and K+)during a hyperosmotic challenge. The hypercapnic response should be mediated by boththe peripheral and central receptors, while the hypoxia and K+ stimulation should involveonly peripheral receptors, which would allow further characterization of hyperosmoticeffects on the different chemoreceptor groups.7MATERIALS AND METHODSSurgical methodsAdult, female Pekin ducks (Anal platyrhynchos), weighing 1.5-3.7 kg, wereobtained from the Animal Care Facility of the University of British Columbia and housedindoors in individual wire cages with free access to food and water. The brachial arteryand the ulnar vein in the wing of each animal were chronically cannulated under localanaesthesia (Xylocaine 2%; Astra, Ont.) with a single piece of polyvinylchloride tubing(Bolab VIII; Bolab, AZ) to form an exteriorized loop. One end of the cannula wasinserted 4 cm into the artery and the other end 5 cm into the vein. Long-term patency wasensured by prior treatment of the cannula with TD-MAC (Polysciences, PA), a heparincompound which binds to the wall of the tubing. The animals were allowed at least 48 hrto recover before the start of any experiments. All experiments were performed inaccordance with Canadian Animal Care guidelines, and approved by the UBC AnimalCare Committee.Physiological responses to acute changes in osmolalityThese experiments were designed to characterize the respiratory, cardiovascularand acid-base effects of osmotic stress.HyperosmolalityExperimental protocol Sixteen ducks were used in these experiments. Each duckwas weighed and placed in a water-cooled, whole body plethysmograph (Fig. 1). Wingsand legs were lightly restrained with filament tape. The head of the animal extended outof the plethysmograph through a dental dam collar and the venous and arterial cannulaswere led out of an air-tight hole. The animals were then left undisturbed for 45 min toallow them to adjust to their surroundings. A 1 hr control period followed in which threearterial blood samples (0.8 ml) were drawn anaerobically for resting blood gas and pHanalysis, using a blood/gas analyzer (IL813; Instrumentation Laboratory, MA) maintainedIwater outflowwaterinflowarterial cannulavenous cannulapneumotachographrectalthermometercannulaeto mass -------'-'■spectrometer-1 IgasinflowlatexcollargasoutflowFig 1.1.^Schematic diagram of the experimental setup for all studies involvingplethysmography. The head compartment was only used when the animal was exposed togases other than room air. See METHODS for details.9at duck body temperature (41°C). The IL813 was calibrated before each sample usingcommercially prepared gas mixtures and pH standards. The remainder of the bloodsample (0.4 ml) was centrifuged and the plasma decanted and immediately frozen.Measurements of resting ventilation and blood pressure (Elcomatic 715A blood pressuretransducer; Harvard Instruments, MA) were taken every 10 min. Heart rate was derivedfrom the blood pressure trace. Eight ducks were each infused with three differentsolutions (4 M NaC1, 2.4 M sucrose, or 0.15 M NaC1) in random order (as outlinedbelow), while the other eight ducks were infused with lactic acid. Every animal was givenat least 48 hr to recover between infusions.Each animal received 7.5 mmol•kg -1 body wt of 4 M sodium chloride (NaC1)infused over 15 min (total volume 5-6 ml). Since this infusion always stimulated saltgland secretion, plasma osmotic pressure was maintained for the duration of theexperiment by subsequent infusion of a 540 mM NaC1 solution at a rate of 0.1 ml•min -1 toapproximately match salt gland secretion. Blood samples were taken at 5, 10 and 15 min,then every 15 min for the remainder of the first hour, and every 30 min in the followinghour. This is termed the infusion period. The total volume of blood removed during eachexperiment was 8.8 ml. Cardiovascular and respiratory variables were measured every 5min for the first hour and every 10 min during the second hour. Each animal alsoreceived two other infusions; an equivalent osmotic load of 2.4 M sucrose overapproximately 25 min (total volume 18-22 ml), and 5-6 ml of isotonic saline (isotoniccontrol infusion). Ducks infused with lactic acid received 0.25 meq•(kg•min) -1 at 0.5ml•min -1 . A maintenance infusion was not required during any of these infusions sincesalt gland secretion was insignificant.Three ducks were infused with the same hyperosmotic NaCl load to measure theresultant change in extracellular fluid volume and hematocrit. Extracellular fluid volumewas measured by injecting 5 ACi of 36C1 in two animals and 22Na in the third animal.After 30 min equilibration at rest, 3 (0.5 ml) arterial blood samples were taken over 1510min and the 4 M NaCl was then infused. Blood samples were taken at 10 min intervalsfor 130 min. The blood was centrifuged and duplicate 100 gl plasma samples weretransferred to 4 ml of Aquasol scintillation cocktail for measurement of 36C1 on a liquidscintillation counter (LS9000, Beckman, CA), while samples were measured for 22Na in aMinaxi 5000 gamma counter (Canberra-Packard, Ont.).Plasma Na+ and K+ concentrations ([Nat], [K+]) were determined by flamephotometry (IL943; Instrumentation Laboratory, MA), plasma Cl- concentration ([Cl-])using a Buehler digital chloridometer (4-2500; Buehler Instruments, NJ) and plasmaosmolality was measured with a vapor pressure osmometer (5500; Wescor, UT). Plasmalactate concentration was measured in two ducks during each of the three treatments in thehyperosmolality series using the method of Noll (1974). The pressure transducer wascalibrated against a mercury manometer. Temperature was monitored in several ducks byrectal thermometer (Physitemp BAT-12; Sensortek, NJ) and did not change throughout theexperiment. Ventilation was measured by whole body plethysmography. Ventilatorychanges in body volume were measured as changes in pressure due to air flow through aFleisch #0 pneumotachograph connected to a port in the plethysmograph. The airflowsignal was measured with a gas pressure transducer (Hewett-Packard 275A) and integratedto yield tidal volume. The system was calibrated with known volumes of air delivered byhand-driven syringe. "CTE was calculated as the product of tidal volume and respiratoryfrequency. All variables were recorded on a Harvard Universal oscillograph writing onrectilinear coordinates.HyposmolalityExperimental protocol The purpose of this experiment was to determine whetherdilution without hyperosmolality or extracellular fluid volume increases would affect acid-base balance. Plasma hyposmolality was induced in five ducks by withdrawing arterialblood with a pulsatile infusion pump which was concurrently replaced with an equal11volume of sterile, deionized water maintained in a water bath at 40°C. Each 10 mlaliquot of blood was immediately centrifuged, the plasma decanted and the red blood cellsreinfused. The control period for measurement of resting variables was identical to thatdescribed for the hyperosmotic experiment above but, during the experimental period,measurements were made and samples taken after the first and each subsequent exchangeof 40 ml of blood. Respiratory, pH and ion measurements were taken as described above,except blood gases were not analyzed. Red cell water content was measured by dryingpacked red blood cells at 105°C to a constant weight.Hyper- and hyposmotic calculations and data analysisData were analysed as the difference from the mean resting value by analysis ofvariance (ANOVA) and Tukeys post-hoc test (Zar, 1984) at a significance level ofP 0.05 using the statistical package Systat (Systat, Evanston, IL). pHa was converted to[H+] for statistical analysis. The bicarbonate concentration was calculated using theHenderson-Hasselbalch equation, assuming a solubility coefficient for CO2 of 0.0282mmol•(L•mm Hg) -1 , and a pK 1 of 6.090 (Helbacka et al., 1964). [SID] was calculated as[SID] = ([Na]+ + [K+]) - ([C1 -] + [lactate]D. All data are shown as differences (mean+ SEM) from the mean resting value established in the 1 hr control period before eachinfusion. Data from the NaC1 and sucrose infusion experiments are compared to theisotonic control infusion experiment.31P Nuclear Magnetic Resonance Spectroscopy31P NMR was performed on twenty-one Pekin ducks in order to quantifyintracellular pH and metabolic changes both systemically and centrally duringhyperosmotic disturbance, and to establish a relationship between pHi and ventilation.Experimental protocol The wing cannula was severed and PVC tubing, longenough to reach the outside of the NMR magnet, was attached to each end. The animal12was placed in a normal sitting position in a cradle and carefully restrained with the aid ofshaped foam pieces and tape. Once inside the magnet, the duck was given 10-15 minutesto relax, and then 3-5 resting spectra were obtained as described in the following sections.During this period, 3 (0.8 ml) arterial blood samples were taken for immediate analysis ofblood gases and pH on a blood/gas analyzer maintained at 41°C. Each duck in whichpectoral muscle was examined was then given an infusion of either hypertonic NaCl (20meq•kg -1), hypertonic sucrose (25.5 meq•kg1), or lactate (9.5 meq•kg -1 ), over 35 min. Infour of the animals, body temperature was monitored via a thermistor inserted into thetrachea. Five spectra were obtained over this period, and at the end of the infusion,recovery was monitored with a further 1-2 spectra. A blood sample was taken in themiddle of each spectrum. Each duck underwent 2-3 random trials spaced at least two daysapart, with no duck receiving the same infusate twice. Ducks that were used for brainstudy were infused only with hypertonic sucrose over 55 min. All solutions were warmedto approximately 40°C just before the start of infusion.31P NMR spectra of pectoral muscleThese experiments were performed in the Department of Chemistry at theUniversity of British Columbia. Each animal (n=16) was placed in a custom-fittedperspex cradle at a slight angle such that the thickest portion of the right pectoral musclelay directly over the surface coil, thus avoiding the sternum. The cradle was then slid intoa 1.89 Tesla horizontal superconducting magnet (Oxford Instruments, Oxford, UK). The31P-NMR spectra were acquired in the Fourier transform mode on a Nicolet 1280spectrometer. The surface coil used was a 3.75 cm, inductively coupled, loop gap (spiralresonating) coil, made from 0.005 cm thick copper foil shielded by Teflon sheets. Thistype of coil was selected because it has a very high 'Q' (approx. 600 unloaded), whichresulted in a very high sensitivity compared to other, capacitatively-coupled, surface coilsthat were tested. However, this coil was not double-tuned, and therefore the magnetic13field was shimmed on a phosphorus sample and not on the proton signal from the animal.The large (approximately 7 cm 3), homogeneous field of the magnet, combined with acradle design that allowed repeated, precise placement of the surface coil and animal inthe magnet, resulted in well resolved spectra with an excellent signal-to-noise ratio (Fig.2). The phosphorus spectra were obtained at 32.5 Mhz, using a 35 Asec pulse width,±2000 Hz sweep width with quadrature detection and 2,048 points per scan. A total of240 free induction decays were collected with a 1.5 sec pulse interval (total time = 7min). Data were zero filled to 4,096 points and the summed free induction decaysmultiplied by an exponential corresponding to 6 Hz line broadening before Fouriertransformation.31P NMR spectra of brainThis study was done in the Department of Radiology and Physiology at theUniversity of Washington, Seattle. The experiments were performed in accordance withthe U.S. Animal Care guidelines and approved by the UW Animal Care Committee. Theprotocol was similar to that used during the pectoral muscle study, except that elevenspectra were obtained over the infusion period, and recovery was monitored with a further6 spectra. The blood/gas analyzer at UW was maintained at 37°C and the data latercorrected to 41°C using a correction factor of -0.015 . °C -1 . Ducks (n=5) were sacrificedat the end of the experiment with an overdose of sodium pentobarbitol.After placement in the cradle and immobilization of the head, the duck was placedin a 2 Tesla, horizontal, superconducting magnet (Oxford Instruments, Oxford, UK). The31P-NMR spectra were obtained in the Fourier transform mode on a GE CSI IIspectrometer with a Nicolet 1280 computer at a frequency of 34.6 Mhz. A 2-turn, 2.5 cmcoil of Teflon-coated copper wire was mounted directly over the head. Since there is nomuscle tissue directly over the cranium in Pekin ducks, the signals monitored by the coilwere from brain tissue up to 1.25 cm deep. Magnetic field homogeneity was shimmed14Fig. 2. Four successive spectra of duck pectoral muscle. Bottom spectrum represents aresting state, with the following 3 spectra obtained 7, 14 and 21 minutes into hypertonicsucrose infusion. Each spectrum is the sum of 240 free induction decays. The scale at thebottom refers to the chemical shift in parts per million with respect to the PCr signal,which is set to 0. Each spectrum has been shifted 1 mm to the right of the spectrumbelow it.toH16with the animal in place using the proton signal from the brain water, so that proton linewidth was in the range of 30-40 Hz. Phosphorus spectra were obtained with a 20 Asecpulse width, ±1500 Hz sweep width, and 2048 data points per scan. A total of 300 freeinduction decays were collected with a 320 msec pulse interval (total time = 5 min), andthe summed free induction decays multiplied by an exponential corresponding to 15 Hzline broadening before Fourier transformation (Fig. 3).31P NMR calculations and data analysisIntracellular pH was calculated from the difference in chemical shifts (6) of the Piand the PCr peaks according to the formula pHi=pK' +log[(L Pi-6A)/(6B-6, Pi)]. Theparameters SA and 6B (the acidic and basic phosphate titration end-points), and pK' (theapparent pK for phosphates), were calculated from Kost (1990) for a body temperature of41°C. The peak area for each metabolite was integrated and corrected for saturation. Thesaturation factor was calculated as the ratio of the individual peak areas obtained with apulse interval of 1.5 sec to the areas obtained with a pulse interval of 18 sec. The latterpulse interval corresponds to approximately five times the longest spin-lattice relaxationtime of the metabolites, the interval required for assumption of complete relaxationbetween scans.To reduce normal inter-animal variation, all data were analyzed using thedifference between the individual's resting value and each experimental value. Data fromthe pectoral muscle experiment were analyzed by ANOVA and Dunnetts post-hoc test, todetermine significant differences from a control value. Control data against which theexperimental data were compared were obtained from three animals that were treated in amanner similar to the experimental animals, except that they were not given an infusion.Data from the brain experiment were analyzed by t-test for paired comparisons. Allresults are presented as mean differences (+ SEM) from resting values, and absolutevalues are indicated where appropriate.17Fig. 3. Representative spectra of duck brain with skull and overlying skin intact. Eachspectrum is the sum of 300 free induction decays. The scale at the bottom refers to thechemical shift in parts per million with respect to the PCr signal, which is set to O. PME,phospomonoesters; PDE, phosphodiesters.1017 —^OZ- 0^ OZ^Itsidd1^1^1^1A,^3141dJ3d , !d30d0d1931P NMR of pectoral muscle during ion-exchange blockadeThe purpose of these experiments was to identify some of the ion-exchange mechanismsinvolved in the pH perturbation during hyperosmolality and to confirm any relationshipbetween pHi and ventilation by manipulation of pHi.Experimental protocol Eighteen female Pekin ducks were used in theseexperiments. Opaque cannulae were used in this experiment to protect the light-sensitiveion-exchange blockers. The animal was fitted with a endotracheal cannula, and thenplaced in a normal sitting position in the cradle and carefully restrained as describedabove. Once inside the magnet, the duck was allowed approximately 15 min to adjust toits surroundings, and then 4-5 resting spectra of the pectoral muscle were obtained. Thespectra were obtained as previously described, except that 172 free inductiondecays/spectra were collected at 4096 points per scan, requiring 6 min per spectra.During this period, 3 (0.8 ml) samples of arterial blood were taken; 0.4 ml were analyzedimmediately for blood gases and pH on a blood/gas analyzer maintained at 41°C, and 0.4ml were centrifuged, the plasma decanted and frozen for subsequent measurement of ions.Plasma Na+ and K+ were measured by atomic absorption spectroscopy (2380; Perkin-Elmer, CT). The endotracheal cannula was attached to a Fleisch #0 pneumotach (totaldead space approximately 3 ml) connected to a differential pressure transducer (DP103-18; Validyne, CA) for measurement of VE for 1 min every 5 min until it was stable.Arterial blood pressure was measured simultaneously.Nine animals were then intravenously infused with physiological saline containingDIDS, 30 Amol•kg -1 in 5 ml, over 10 min. Six other ducks received 20 mg•kg -1 amiloride(0.075 Amol•kg -1) in the same manner. Blood samples were taken and N:TE and bloodpressure monitored 20 and 30 min after the end of infusion to ensure that most variableswere stable. One NMR spectrum was acquired over the last 6 min of the equilibrationperiod. The animals were then infused with 26.5 mmol•kg - ' body wt. sucrose (approx. 2520ml) over 36 min. All variables were monitored every 6 min during the infusion periodand during the 12 min recovery period.Three animals were infused only with the sucrose, and blood samples were takenafter 15 and 30 min of infusion for measurement of total CO2 using a Carle AGC series100 gas chromatograph. Calibration was performed using 25 Al of 20 mM NaHCO3standards.Calculations and data analysisCalculations were performed as previously described. All results were analyzedusing paired-comparison ANOVA.Responses to respiratory stimuliThese experiments were performed to test the hypothesis that the intracellular acid-base perturbation caused by acute hyperosmolality would depress the normal ventilatoryresponse to a respiratory stimulus that acted through acidotic pH reception (CO2), whilethe response to stimuli not known to involve pH changes (02, K+) would be unaffected.Experimental protocol Eleven Pekin ducks were exposed to one or more knownrespiratory stimuli: hypoxia (10% 02, 90% N2), hypercapnia (3.5% CO2 in air) or an i.v.infusion (approx. 0.5 ml) of a K+ load equivalent to 50% of the calculated extracellular[K+J. Seven ducks were exposed to both hypoxia and hypercapnia, one duck to hypoxiaonly, and three ducks were given two different K+ loads. All ducks were allowed at leastfive days between trials. Each duck was placed in a whole body plethysmograph, with theneck extending out of the main chamber through a double layered dental dam collar into aseparate head chamber (Fig. 1). The head chamber was ventilated with room air at 51•min-1 . An opening at the top of the body plethysmograph was fitted with a Fleisch #0pneumotach for measurement of VE. The body compartment was surrounded by acirculating cold water bath which maintained duck body temperature at 41°C.21The animal was allowed 20 min to adjust to its surroundings, and then ventilationwas recorded for 1 min every 10 min until it was stable (usually 4 recordings). Threesamples of 0.7 ml of arterial blood were taken, with 0.4 ml analyzed for blood gases andpH on a blood/gas analyzer maintained at 41°C, while 0.3 ml were centrifuged, theplasma decanted and immediately frozen. The inflowing air and the end tidal gases (FET)were monitored via polyethylene 60 cannulae connected to a MGA 200 clinical massspectrometer (Centronic, U.K.) and sampled concurrently with VE. Arterial bloodpressure was also recorded and heart rate was derived from the blood pressure trace.Immediately after the resting control period, the animal was exposed to either thehypercapnic or the hypoxic gas mixture for 2 min by turning a stopcock connected to thehead chamber, which allowed the premixed gas to flow through the chamber at 5 1. min -1 .Three of the animals were maintained on room air but infused with a K+ bolus (approx.volume 0.5 ml) over 30 sec. Ventilation, blood pressure, heart rate, end tidal gases andthe inspired gases were continuously monitored. A blood sample was collected at 2 min.The animal was allowed 15 min to recover while breathing room air, with all variablesbeing sampled at 10 and 15 min post stimulus to confirm a return to resting levels. Eachanimal was then infused with 26.5 mmol•kg body wt. -1 sucrose (approx. 25 ml) over 40min. All variables were monitored every 7.5 min until 22.5 min, at which time therespiratory stimulus was repeated. A final recording was made 15 min after the end ofthis second respiratory challenge.Three ducks were twice infused with K+, once as a bolus, and once with the sameK+ load mixed in with the sucrose to assess the effect of rate of extracellular K+ increaseduring dilution acidosis. These same animals were also given a bolus of 150 mM NaCl asan experimental control. These trials were given in random order approximately 5 daysapart. The K+ load was calculated to be 50% of extracellular K+ assuming anextracellular space of 25% body weight (Ruch and Hughes, 1975) and a mean plasma[K+] of 2.5 meq•kg -1 . This load was empirically decided upon in preliminary experiments22because it resulted in a clear, brief ventilatory increase in normosmotic animals withoutcausing excessive stress (as measured by changes in cardiovascular variables). Changes intotal extracellular K+ caused by the sucrose infusion were taken into account by assumingan increase in extracellular fluid volume of 10% and a constant plasma [K+] of 2.5meq. kg-1 .The mass spectrometer was calibrated with pure N2 gas and a precision analyzedmixture of 14.00% 02, 8.19% CO2 , 8.24% argon with the balance N2. Hypercapnic orhypoxic gas mixtures were mixed using a flow meter, and the level of CO2 or 02 checkedby the mass spectrometer.Calculations and data analysisAll results were analyzed by random block two-way ANOVA. With somevariables, high inter-animal variability necessitated re-analysis by paired-comparison, two-way ANOVA. Since the recovery measurements at 10 and 15 min post stimulus were notsignificantly different for any measured variable, the two values were averaged forstatistical analysis. Results are reported as mean + SEM, or mean difference + SEM.Significant differences are from resting, normosmotic levels unless otherwise stated.Significant differences from hyperosmotic values always refers to the measured value at22.5 min into sucrose infusion. 02 consumption was calculated as N .702 = VE • [1 -(FE02 FECO2) / 1 - (FIO2 FICO2)] • FIO2 - VE • FEO2, where FE and F1 representthe fractions of end tidal expired and inspired gas, respectively.2 3RESULTSPhysiological responses to acute changes in osmolalityHyperosmolality versus lactateInfusion of 4 M NaC1 caused a significant increase in mean extracellular fluidvolume of 10±5%. The expansion was complete 30 min after the start of infusion andwas maintained for the rest of the sampling period. This extracellular fluid volumeexpansion must have been due to the movement of intracellular water since the volume ofthe NaC1 infusate never exceeded 1% of extracellular fluid volume. Resting hematocritwas 39+4% and decreased 4±2% with the infusion.There were no significant changes in plasma osmolality, [Nat] or [C11 duringthe isotonic control infusion (Fig. 4). Plasma ions were not measured during lactateinfusions. Plasma osmolality increased by a maximum of 31+1 and 27±2 mosm•kg -1above resting during the NaC1 and sucrose infusions and remained significantly higherthroughout the infusion period (Fig. 4). Plasma [Nal increased and decreased amaximum of 15+4 and 16+2 mec•lcg -1 with the NaC1 and sucrose infusions, respectively(P <0.05). However, plasma [C11 increased a maximum of 21+1 meq . kg -1 and decreaseda maximum of 11±2 mec•lcg -i with the NaC1 and sucrose infusions, respectively(P <0.05). The increase in plasma [Cl-] remained significantly higher (5-6 meq•kg -1) thanthe increase in plasma [Nat] throughout the NaC1 infusion period and during the first hourof the sucrose infusion period. [SID] decreased significantly during the hypertonicinfusions, primarily because of the relative increase in the [Cr] (Table I). Both plasma[Na] and [C11 returned to resting levels by the end of the sucrose infusion period.Plasma [K+] remained constant for the first 90 min and then increased slowly butsignificantly over the last 30 min during all three infusion periods (Fig. 4). Plasma[lactate] only changed significantly during the lactate infusion, increasing by 7.8+0.4mmol•kg1 (Table I).2 4Fig. 4. Changes observed in osmotic pressure, plasma [Nal, [cr], and [Kl duringcontrol (isotonic saline), 4 M NaC1 and 2.4 M sucrose infusions. Mean (±SEM)differences from resting values are shown. n=8.• =control, ^ =NaC1, A =sucrose infusion.* indicated point significantly different from rest (P <0.05)** significantly different from rest throughout infusion30 -50 -:^•^CD^30 -co20 -CD'a(,)E^lo00 -0 -^0-10 —•- - --A- .......... D....105 120 1350^15^30^45^60^75^90Time (min)-30 10^15^30^45^60^75^90^105 120 135105 120 1350^15^30^45^60^75^90Time (min)0^15^30^45^60^75^90^105 120 13520a)+ 0coco^—EcoEL^-20*1.50.50 .0-0.5-1.0-2040 -26Table I. Ionic, pHa and blood gas changes 30 min after the start of hyperosmoticinfusion. Data are shown as the difference from resting values (i.e. when t=0). Absoluteresting values are in parentheses; mean+SEM.IsotonicsalineHypertonicNaC1Hypertonicsucrose Lactate0.0± 1.4 29.5+2.6* 27.0+2.5*(280+6) (282+4) (283+5)1.0±1.6 15.0+4.1* -16.0+2.0*(147+2) (148+5) (148+4)0.2 + 0.2 -0.4+0.2 -0.1+0.2(2.4±0.4) (2.6+0.6) (2.5+0.4)0.0+0.5 20.0+1.2* -11.0+2.1*(108+2) (108+4) (111+3)-0.1a 0.8a 0.2a 7.8+0.4*(2.0) (2.5) (2.0) (2.1+0.8)1.9±4.1 -6.5±7.4* -5.1±2.7*(43.0+4.0) (42.0+ 10.0) (41.0+5.0)0.00+0.00 -0.06+0.01* -0.05+0.01* -0.08+0.01*(7.48+0.03) (7.48+0.03) (7.49+0.03) (7.48+0.02)0.5±0.7 2.0+2.0 3.0+1.6 -3.9+0.6*(29+4) (31+3) (31+3) (29.7+0.8)0.0+1.4 0.9+1.0 -2.3±1.6 11.6+1.8*(84.1+4.4) (87.3+7.9) (86.1+2.4) (89.6+3.3)-0.4+0.5 1.5+1.1 0.5+0.6 5.5+0.4(20.8+2.4) (22.3+2.0) (21.9+2.3) (20.5+1.0)isotonic saline (control)OP(mosm•kg")Na(meq . kg")K+(meq•kg-1)cr(meq.kg")Lactate-(mmol . kg-1)SIDpHaPaCO2(torr)Pa02(torn)HCO3-(mmol.kg")n=8; a n=2* significantly different from27With both hypertonic infusions, only minor changes in respiratory variables wereobserved despite significant decreases in pHa (Fig. 5). During the NaC1 and sucroseinfusion experiments, VE did not change significantly from resting VE except for atransient 53 % increase 15 min after the start of the NaC1 infusion, due to small increasesin both respiratory frequency and tidal volume. However, pHa decreased significantlywithin 5 min after the start of both the NaC1 and sucrose infusions. In one animal, "CrEdecreased 45% by the end of the sucrose infusion and remained low for 90 min, despite afall in pHa of 0.15 pH units and a rise in Paco2 from 34 to 50 torr (Fig. 6). Thereduction in VE was almost entirely due to a decrease in respiratory frequency. Similarbut less dramatic changes also occurred in two other sucrose infused animals. The pHacontinued to decline throughout the NaCl experiment, but pHa began to recover 60 minafter the start of the sucrose infusion (Fig. 5). Neither VE or pHa changed significantlyover the isotonic control infusion period. With the exception of the three sucrose infusedanimals mentioned above, there was no change in the pattern of respiration, i.e. in tidalvolume or respiratory frequency, during any of the infusions. There were no significantdifferences in PaCO2 or Pa02 among the three treatments at any time (Fig. 5). PaCO2increased significantly an average of 3.5+1.5 torr and Pa02 decreased significantly anaverage of 4±2 torr during all three infusion periods. Calculated bicarbonateconcentration did not change significantly over the infusion periods. There were nosignificant changes in either mean arterial blood pressure or heart rate during any of theinfusion.Lactate infusion significantly decreased pHa from resting levels, and the change inpHa was not significantly different from that due to the hypertonic infusions.increased significantly, and this was reflected in an increase in Pa02 of 13.4+1.7 ton anda decrease in PaCO2 of 4.3+0.5 ton (all P <0.05) (Table I).28Fig. 5. The changes observed in pHa, 'TE, PaCO2 and Pa0 2 during control (isotonicsaline), 4M NaC1, 2.4M sucrose and lactate infusions. Mean (±SEM) differences fromresting values are shown. n = 8.• =control, 0 =NaC1, A =sucrose, • =lactate infusion.* indicated point significantly different from rest (P <0.05)** significantly different from rest throughout infusion# significantly different from 0 for all three treatments0.02 -0 .00 -1-0.02 --0.04-0.06 - ...-0.08 --0.10 --0.12^1^I^I^I^I^I^I^I^IO 15^30^45^60^75^90^105 120 135^•^O 15^30^45^60^75^90^105 120 135161412108-z-8 64-442CO^00-2-4-6-8-10105 120 135 105 120 135400300200100• --100-200-300O 15^30^45^60^75^90Time (min)O 15^30^45^60^75^90Time (min)30Fig. 6. The changes in pHa, VE, and PaCO2 measured in a single animal during infusionof 2.4 M sucrose (7.5 mmol•kg-1).7.50 —7.457.40 —7.35 —7.30 —1^1^I^I^I^I^I^I0^15^30^45^60^75^90^105^120Time (min)55 —50 —35 —I^I^I^I^I^I^I^I300^15^30^45^60^75^90^105^120Time (min)1250 —1000 —750 —500 —250 —0 I^I^I I^I^I I I0^15^30^45^60^75^90^105^120Time (min)EEE7.253 2HyposmolalityReplacement of plasma with deionized water did not elicit the changes in acid-basestatus caused by hyperosmolality. There was a significant decrease in plasma osmolality of21+1 mosm•kg -1 from the average resting osmolality of 282+3 mosm•kg -1 . Plasma[Na] decreased 8+2 meq•kg -1 while plasma [CI] decreased 6+2 mec•kg - ' (P <0.05).There was no significant difference between the changes in plasma [Nal and [Cl -] at theend of the experiment. There was no significant effect of hyposmolality on pHa,hematocrit, [Kl, red cell water content, or mean arterial blood pressure. Heart rateincreased 60% by the end of the experimental period, but this change was not significant.31P nuclear magnetic resonance spectroscopyPectoral muscleResting, mean body temperature was 41.0±0.5°C, and it did not vary by morethan 0.2°C during any of the experiments. There was no significant difference in anymeasured variable between the sucrose or NaC1 infused animals, therefore the data for thetwo groups were combined and analyzed together (results shown in Fig. 7). pHadecreased 0.08+0.01 units from a resting value of 7.52+0.04 over 35 min of hypertonicinfusion (P < 0.05), and remained significantly different from rest 15 min post-infusion.PaCO2 rose 3.3+0.8 torr during the infusions (P <0.05), and returned to resting valuesafter 15 min of recovery. Pa02 did not change significantly over the infusion period, butdecreased 5+5 ton upon cessation of the infusion and remained low during the recoveryperiod (P < 0.05). The large decrease in Pa02 at 7 min post-infusion was largely due toone animal. Mean resting muscle pHi was 7.13±0.04 and it increased significantly by0.11+0.02 units within the first 7 min of perfusion and remained relatively stable over theinfusion period. The maximum increase in muscle pHi in individual animals ranged from0.059 to 0.316 pH units. Post-infusion recovery was very fast, with pHi back at restingvalues within 7 min, and slightly overcompensated at 15 min. There was no significant3 3Fig. 7. Changes in pH and blood gases during hypertonic infusions (solid line) and lacticacid infusion (dotted line). Filled symbols represent the infusion period.14121086420-2-4-6-8-10-^7^14^21^28^35^42^49^0^7^14^21^28^35^42^490^7^14^21^28^35^42^49^0^7^14^21^28^35^42^49Time (min) Time (min)3 5change in the relative concentrations of any of the high energy phosphate metabolitesmeasured. There was a trend toward a relative increase in PCr, and a trend toward arelative decrease in Pi during infusion, neither of which were significant.The time course and total decrease in pHa resulting from lactic acid infusion werenot significantly different from the acidosis measured during the hypertonic infusions (Fig.7), except that recovery from the lactacidosis following the cessation of infusion was morerapid, and was complete in 15 min (Fig. 7). PaCO2 decreased 5.2+1.6 torn during theinfusion period (P <0.05), but returned to resting levels 15 min post-infusion. Pa02 rose12+1 ton during the infusion (P <0.05) and recovered in 15 min, slightly overshootingresting values. Muscle pHi decreased significantly by 0.18+0.06 pH units within 15 minof infusion and then partially recovered by 30 min and remained approximately 0.06 pHunits below resting values for the duration of the infusion. The maximum decrease inmuscle pHi in individual animals over the course of the infusion ranged from 0.114 to0.370 pH units. Recovery was essentially complete by 15 min post-infusion. The restingmuscle pHi, 7.16+0.03, was not significantly different from the resting pHi during thehypertonic experiments. There was no significant change in any of the measuredphosphate metabolites during the infusion.BrainInfusion of hypertonic sucrose decreased pHa 0.186+0.07 units after 47.5 min ofinfusion (p < 0.05) from a resting pHa of 7.47±0.05 (Fig. 8). During the recoveryperiod, pHa decreased even further to 0.32+0.07 units below the resting level. Therewas no significant change in PaCO2 except for an increase at 17 min into the infusion.Considerable inter-animal variation was observed, with the PaCO2 in one animal risingfrom 33.5 to 59.4 ton by the end of the infusion, peaking at 66.4 ton midway through therecovery period and then declining (data not shown). Pa0 2 remained unchanged duringthese experiments. Mean resting brain pHi was 7.08+0.02. Brain pHi did not show a360^10^20^30^40^50^60^70^80^90Time (min)Fig. 8. Changes in arterial pH and brain pH during hypertonic sucrose infusion.Filled symbols represent the infusion period.3 7consistently significant change during the experiment, and was significantly increased onlyat 30, 35, and 50 min into the infusion period (Fig. 8). Mean brain pHi during theinfusion was 7.11+0.01.Ion-exchange blockadeDIDS Infusion of DIDS did not significantly affect pHa or muscle pHi (Fig. 9).Resting pHa was 7.48+0.01 and resting muscle pHi was 7.14±0.01. However,increased slightly upon DIDS infusion from a resting level of 326±57 to 428±73m1•(kg•min) -1 (P <0.05) (Fig. 10). Neither PaCO2, Pa0 2 , respiratory frequency, tidalvolume, plasma [Nal, [C11 nor [K -4 ] changed significantly. Mean arterial bloodpressure decreased 9+4 mmHg from a resting level of 230+5 mmHg and heart rateincreased 41±8 beats•min -1 (P <0.05).Sucrose infusion caused a significant decrease in pHa to 7.39+0.02, and decreasedmuscle pHi to 7.06+0.03 (P <0.05) (Fig. 9). The decrease in pHa began immediatelyupon infusion, but muscle pHi was not significantly affected until 21 min into theinfusion. VE continued to increase throughout the infusion (P <0.05), rising to 712+157ml•(kg•min) -1 (Fig. 10). PaCO2 did not change at any time, but Pa02 rose significantlyfrom 79±2 to 91+2 torr (Fig. 9). Mean arterial blood pressure continued to decrease to212+6 mmHg, and heart rate increased a total of 62+19 beats min -1 (P <0.05). Plasma[Cl-] decreased significantly from 108+2 to 98+2, and plasma [Na] fell from 125+2 to113+4 meq•kg -1 by the end of the infusion (Fig. 11). Plasma [KI also significantlydecreased during the infusion. All variables except plasma [K+] remained significantlydifferent from resting levels during the recovery period.Amiloride Infusion of amiloride significantly decreased pHa and muscle pHi (Fig.12), and increased 1:TE (Fig. 10) (P < 0.05). The resting levels of all three variables werenot significantly different from those measured in the DIDS-treated animals. PaCO2decreased 7.1+1.1 ton from a resting value of 28.5+2 ton, while Pa02 increased 10±23 8Fig. 9. The effect of DIDS treatment on the pH and blood gas response tohyperosmolality. Open symbols represent animals given a sucrose load only, and are datafrom the 31P NMR experiment on pectoral muscle shown for comparison purposes.Closed symbols represent animals given an ion-exchange blocker. The first closed symbolrepresents the response to the drug alone, and the last 2 closed symbols represent therecovery period.* significantly different from resting (P <0.05).55^61 67 73 79 8510 —-20 —I^I73 79 8573 79 85-12 /^//1^37 43 49 55 61^67Time (min)549^-37 43 5^5 61^67Time (min)0.20 —0.15 —0.10 —0.05 —0.00 —-0.05 —-0.10 —-0.15 /1^37 43 4967 8573 7949^55 61-0.125 /^// 1^37^43I0_0.025 —0.000 —-0.025 —Ic0-0.050 —-0.075 —-0.100 —6 —4 —2 —-2 —-4 —6 —c.104coa_0-1 6 —12 —O--4 —-8 —600 -40DIDS500 -* *^_400 -300 - 200 -100 -0 ^1 /// 1^1^1^1^1^1^I^I1^37^43^49^55^61^67^73^79^85400 -Amiloride1^37^43 49^55 61^67^73^79^85Time (min)Fig. 10. Changes in ventilation in response to hyperosmolality during ionexchange blockade. The first symbol represents the response to the blockeralone, and the last two symbols represent the recovery period. * significantlydifferent from resting.//^ I37 43 49 55 61 67 73 79 85coa-a)E1050———■ -5 —-1-ca -10 —Z< -15 —•^ -20 —5< -25 —-30 1110 ---,iii.._ 5 —-8-a) 0 —E-5 —■7--co-10 —Z -15 —-20 —•-25 —Z3a -30 —-35 1141DIDS// I^I^1^I^1^1^I^I37 43 49 55 61 67 73 79 85Amiloride*Time (min)Fig. 12. Changes in plasma [Nal and [C1] during hyperosmolality and ionexchange blockade. The first symbol represents the response to the blockeralone, and the last two symbols represent the recovery period. * points fromboth lines significantly different from resting. # amiloride only significantlydifferent from resting.4 2Fig. 12. The effect of amiloride treatment on the pH and blood gas response tohyperosmolality. Open symbols represent animals given a sucrose load only, and are datafrom the 31P NMR experiment on pectoral muscle shown for comparison purposes.Closed symbols represent animals given an ion-exchange blocker. The first closed symbolrepresents the response to the drug alone, and the last 2 closed symbols represent therecovery period.* significantly different from resting (P < 0.05).0.025 —-0.000 —-0.025 —c0-0.050 —o.-0.075 —-0.100 —-0.125'1-0.150^/1 370.20 —0.15 —0.10 —0.05 —0.00 —Q. -0.05 —-0.10-0.15 —-0.20 —I^I^I -0.25 i //^ I^I8555 61^87 73 798543 1^37 43 4961^67 73 7949 557349^55 61^87Time (min)79 8528 —24 —20 —1612 —8-4 —o-4 —-8 —73 79 8537 43 49 55 61 6708 —6 —42 —O 0 —cr.1o_ -2 —-4 —-6 —-8 —-10^///1^37^43Time (min)44torr from 87+2 ton (Fig. 12) (P <0.05). Plasma [C11 increased 6±2 meq•kg -1(P < 0.05), while [Na] also increased 6+4 meq•kg - ', although this change was notsignificant (Fig. 11). There was no change in plasma [Kl, blood pressure or heart rate.Sucrose infusion decreased pHa further by 0.11+0.02 pH units (Fig. 12). MusclepHi remained low (0.16+0.04 below resting) for 21 min of sucrose infusion (P < 0.05),but then recovered to a value (0.09+0.06 pH units below resting) not significantlydifferent from the resting value. VE remained high (P <0.05) and continued to increaseduring the sucrose infusion to 603+53 ml•(kg•min) -1 , except for a transient return toresting levels 9 min into the infusion (Fig. 10). The increase in VE was primarily due to a55% increase in tidal volume. PaCO2 increased during the infusion to a maximum of26+2 ton, but remained significantly lower than resting values (Fig. 12). Pa02 increasedfurther during the infusion to 106+2 ton (P <0.05). Plasma [C11 decreased 21+3meq•kg-1 (P < 0.05) (Fig. 11), a significantly greater decrease than during DIDStreatment. Plasma [Nal fell 24±9 meq•kg -1 and plasma [K1 also decreased significantlyduring the sucrose infusion, but both ions recovered to levels not significantly differentfrom resting levels by the end of the infusion. The drop in [K1 was not significantlydifferent from that during DIDS treatment, and averaged 0.7 mEq. Mean arterial bloodpressure decreased transiently by 24±7 mmHg midway through the infusion, and heartrate increased slowly to reach 43+15 beats . min -1 over resting levels by the end of theinfusion (P <0.05). VE, pHa, PaCO2, Pa02 , and plasma [Cl -] remained significantlydifferent from resting values at the end of the recovery period.Total CO2 Total CO2 decreased significantly during sucrose infusion, from acontrol normosmotic level of 24.5+0.1 mM to 21.8+0.5 mM after 30 min of infusion(P < 0.025).45Responses to respiratory stimuliHypercapnia Inhalation of 3.5% CO2 decreased pHa from 7.48+0.01 to7.44±0.01 in 2 min (P <0.01) (Fig. 13). VE increased from 321+36 to 458+34m1•(min•kg) -1 in 1 min and to 656+61 ml•(min•kg) -1 in 2 min (P <0.001). The increasein VE was due to an increase in tidal volume as respiratory frequency did not change.PaCO2 increased from 31.5+1.2 to 38.1+1.2 torr, while PaO2 increased from 95+2 to105+2 torr in 2 min (P <0.001) (Fig. 13). While FETCO2 did not vary, FETO2 increasedsignificantly from 15.3+0.2% to 17.3+0.2%. Both FETCO2 and FETO2 closely reflectedchanges in blood gases throughout the experiment. TO2 rose from 19.0+2.5 to27.8+4.5 ml 02' (kg.min) -1 (P < 0.025) (Fig. 14). None of these variables weresignificantly different from resting values after 15 min of recovery. Heart rate, plasmaions, and plasma lactate were not significantly affected by the period of hypercapnia.Mean arterial blood pressure and plasma [K+] did not vary at any time over theexperiment.Sucrose infusion decreased pHa 0.06+0.01 units (P < 0.01), while PaCO2increased 3.9±0.8 torr (P < 0.001) and VE and Pa0 2 did not change over the infusionperiod (Fig. 13). 702 increased to 25.2+4.9 ml 02•(kg•min) -1 (P <0.025) (Fig. 14).FETCO2 rose slightly, while FETO2 decreased slightly during the infusion (P <0.05).Heart rate increased significantly from 129+8 to 196±19 beats min -1 . After 22.5 min ofinfusion, plasma osmolality increased from 284+5 to 319+6 mosm•kg -1 , plasma [Nat]decreased 19+1 meq•kg -1 from a normosmotic value of 144+2 meq•kg -1 , and [Cl-]decreased 12±1 meq•kg -1 from a normosmotic level of 108+5 meq•kg -1 (all P <0.05).Hypercapnia during the sucrose infusion further decreased pHa from 7.42+0.01(the value after 22.5 min of sucrose infusion) to 7.36+0.02 in 2 min (P < 0.01), while VEincreased from 387+75 to 566+87 ml.(kg•min) -1 in 1 min and to 799+124 in 2 min(P <0.01) (Fig. 13). The ventilatory responses to CO2 in normosmotic and hyperosmoticducks were not significantly different. However, there was an increase in the PaCO24 6Fig. 13. pHa, blood gas and ventilatory changes in response to hypercapnia and sucrose.Black bars represent the period of hypercapnia. The dotted line indicates the beginning ofthe sucrose infusion. * significantly different from resting value. # significant differencebetween 1st and 2nd hypercapnic period.IS #I^I^I^I^1^I^1^1^1^1-5 0^5 10 15 20 25 30 35 40 45 50 557.550 -7.525 -7.500 -7.475 -7.450 -7.425 -7.400 -7.375 -7.350 -7.325900 -800 -700 -600 -CD 500 -C400300Lu^200 -100 -0111^11^I I^111-5 0^5 10 15 20 25 30 35 40 45 50 55Time (min)50 -45 -8^400^35 -Ucoa_.#30 -25 -1111111^1^111-5 0^5 10 15 20 25 30 35 40 45 50 5520I^I^1^I^I^I^I^1^1^1-5 0^5 10 15 20 25 30 35 40 45 50 55Time (min)480^5^10^15^20^25^30^35^40^45^50^55Time (min)Fig. 14. Changes in V42 associated with hypercapnia•and hypoxiao before andduring hyperosmotic stress. Black bars represent the period of hypercapnia or hypoxia.The dotted line indicates the beginning of the sucrose infusion. * significantly differentfrom rest.4 9during hyperosmolality which was associated with the same VE as in normosmoticanimals, and there was a corresponding decrease in the magnitude of the ventilatoryresponse to any given increase in PaCO2 in hyperosmotic ducks. This is shown by theright-shifted curve of hyperosmotic animals (Fig. 15). There was, however, no change insensitivity to PaCO2 as measured by the slopes of the lines (Fig. 15). PaCO 2 increasedfrom 35.5+1.2 to 47.4±2.5 torr in 2 min (P < 0.001), a significantly greater change thanthe change in PaCO2 during the normosmotic hypercapnia (Fig. 13). Pa02 increasedfrom 94+3 to 102+2 (P <0.025). FETCO2 did not change with respect to the FETCO2during the sucrose infusion, but was significantly increased over normosmotic FETCO2 ,and FETO2 increased significantly from 14.7±0.3 to 16.6+0.3%. TO2 was significantlygreater than resting, normosmotic levels, but did not increase significantly from thehyperosmotic level (Fig. 14). Heart rate and plasma electrolytes did not changesignificantly over the 2 min test.The pHa remained low (7.39+0.01) 10 min post hypercapnia, while 'CIE, Pa02(Fig. 13), and FETO2 returned to values not significantly different from normosmotic orhyperosmotic values. PaCO2 , FETCO2, and TO2 remained significantly higher thannormosmotic levels but not hyperosmotic levels. Heart rate remained significantlyelevated over resting normosmotic levels by 94+30 beats min-1 after 10 min recoveryfrom hypercapnia. Plasma osmolality was further increased to 338+6 mosm•kg -1 , a totalincrease of 54 mosm•kg -1 from resting osmolality. Plasma [Nat] decreased to 122±1meq•kg-1 , a total of 21+1 meq•kg -1 lower than the normosmotic level, and [Cl -] decreasedto 95+4 meq•kg -1 , a total of 13+2 mec•kg - i below the normosmotic level (all P <0.05).Plasma [lactate], which did not increase significantly during the sucrose infusion(1.7+0.4 to 3.0+0.4 mmol•kg - '), was significantly elevated 10 min post hypercapnia(3.9+0.4 mmol•kg -1 ).Hypoxia Arterial pH was increased from 7.48±0.01 to 7.54+0.01 units(P <0.001) after 2 min of breathing 10% 02, but returned to resting levels after 15 min ofI^1^I^1^I^r40^50^60^70I^'^I90 10080501000 ^900 -800 -700 -600 -500 -400 -300 -200 -100 ^251 000900 —800 —700 —600 —500 —400 —300 —20030I^V^I^I^I30 35 40PaCO2 (torr)45 50 55Pa02 (torr)Fig. 15. CO2 and 02 ventilatory response curves. Dashed line represents thecontrol ventilatory response to the gases, and the solid line represents theventilatory response to the gases during hyperosmolality.51recovery (Fig. 16). 'CTE increased significantly from 261+19 ml•(min•kg) -1 to 414+40ml•(min•kg) -1 in 1 min and to 519+39 ml•(min•kg) -1 in 2 min, and returned to restinglevels within 15 min (Fig. 16). The increase in VE was primarily due to a 33% increasein respiratory frequency during the first minute, which then remained stable, so that in thesecond minute, the increase in VE was mostly due to a 46% increase in tidal volume.PaCO2 decreased from 30.6+0.6 to 27.5+1.6 ton . (P < 0.05), and Pa02 decreased from91+1 to 45+2 ton (P <0.001) after 2 min of hypoxia (Fig. 16). FETCO2 declined from5.3+0.2 to 4.5+3.1 % (P <0.001) after 2 min, while FETO2 decreased from 15.2+0.5to 7.2+0.4% (P <0.001). Heart rate increased by 16+7 beats . min-1 from a restingnormosmotic level of 124+11 beats. min -1 (P < 0.05), but VO2 (Fig. 14), plasma ions, andplasma [lactate] were not affected by hypoxia. All variables returned to resting levelswithin 15 min except Pa02, which was depressed by 3.0+1.1 ton. Mean arterial bloodpressure was not affected at any time during the experiment.Sucrose infusion caused a 0.06+0.01 unit decrease in pHa (P <0.01), with nosignificant increase in VE or Pa02 (Fig. 16). PaCO2 increased from 31.2+0.9 to35.4+1.9 ton. (P < 0.001). FETCO2 rose slightly (P <0.05) while FETO2 decreased(P <0.025). /132 increased significantly from a resting normosmotic level of 16.1+1.2ml 02•(kg•min) -1 to a maximum of 20.6+1.9 ml 02•(kg•min) -1 (Fig. 14). Heart rate rosefrom 123+11 to 169+21 beats min -1 (P <0.05). Plasma osmolality increased 31+3mosm•kg -1 after 22.5 min of infusion from a normosmotic level of 287+3 mosm•kg -1 ,[Na] decreased 18+1 meq•kg -1 from a value of 144±1 meq•kg -1 , and [Cl-] decreased10+1 meq•kg -1 from a normosmotic concentration of 106+1 meq•kg -1 (all P <0.05).Plasma [Kl and [lactate] remained constant throughout the experiment.In the second hypoxic period, pHa significantly increased back to restingnormosmotic levels. This pH change was identical to that in the first hypoxic episode.The increase in VE, though, was double that of the increase during the control hypoxia(Fig. 16). This is reflected in a 7.3+1.4 ton decrease in PaCO2 from the hyperosmotic52Fig. 16. pHa, blood gas and ventilatory changes in response to hypoxia and sucrose.Black bars represent the period of hypoxia. The dotted line indicates the beginning of thesucrose infusion. * significantly different from resting value. # significant differencebetween 1st and 2nd hypoxic period.50 —45 —40 —O•FPN 35 —0(_)C1:3^30 —CI_25 —111111^f-5 0^5 10 15 20 25 30 35 40 45 50 5520•#7.550 —7.525 —7.500 —7.475 —(a 7.450 —7.425 —7.400 —7.375 —7.350 —7.325^1^1^1^1^I^1^1^I^1^II^1-5 0^5 10 15 20 25 30 35 40 45 50 55100 —90 —80 —70 —0^60 —t0050 —40 — 900 —800 —700 —600 —500 —400 —300 —200 —100 —*##*#CEE0 1^1^1^I^I^1^I^1^I^1^1^1 30-5 0^5 10 15 20 25 30 35 40 45 50 55Time (min) Time (min)-5 0^5 10 15 20 25 30 35 40 45 50 5554level (P <0.001), a significantly larger decrease compared to the decrease during thenormosmotic hypoxia. However, the fall in Pa0 2 to 41.8±2.1 torr) was not significantlydifferent from that in normosmotic ducks during hypoxia. Thus, there was an increase inthe magnitude of the ventilatory response relative to the decrease in Pa02, shown by theupward-shifted point representing hyperosmotic hypoxic animals (Fig. 15). The increasein VE was due to a 63% increase in tidal volume in the first minute, which rose to 95% inthe second minute, a significantly larger response than during the normosmotic hypoxia.Respiratory frequency rose 28% the first minute and 42% the second minute, a responsenot significantly different from the hypoxic response in normosmotic ducks. While FETO2decreased to exactly the same level as in the first hypoxic bout (P <0.001), FETCO2 and70D2 were not affected. Heart rate increased significantly from 122+9 to 215+22beats . min-1 after 2 min of hypoxia, and plasma [lactate] increased from 2.3+0.3mmol•kg-1 to 3.4+0.4 mmol . kg-1 (P <0.05).At 10 min post hypoxia, YE, PaCO2, Pa02, FETCO2, FETO2 and .i'702 hadreturned to levels not significantly different from either resting normosmotic orhyperosmotic levels, while heart rate and plasma lactate levels remained significantlyhigher than resting normosmotic values but not the hyperosmotic values. pHa had furtherdeclined to 7.40±0.02 (P <0.01), osmotic pressure had increased 55+4 mosm•kg -1 , finalplasma [Na} had decreased 21+1, and plasma [Cl -] 13+1 meq . kg -1 (all P <0.05).Potassium A K+ bolus in resting, normosmotic animals increased tidal volumesignificantly for 3 breaths, at which point it returned to a volume not significantlydifferent from resting levels (Fig. 17). The K+ bolus after 22.5 min of sucrose infusionincreased tidal volume for more than 15 breaths (P <0.01) and did not return tohyperosmotic levels (which were not significantly different from rest) for over 2 min (Fig.17). The K+ bolus did not significantly affect inter-breath interval, heart rate or meanarterial blood pressure, while neither K+ added to the sucrose infusion nor the bolus of55150 mM NaC1 had any significant effects on respiratory or cardiovascular variables at anytime.*min Breaths140 - ***6-- -0%.120 -100 -80 -40 -20 -60 -o —I^'^I^I^I^I^'^I^'^I^I^'^I5^15 30 1^2^3^4^5^6^7^8^9^10 11 12 13 14 15-40A-20 -56Fig. 17. The respiratory response to bolus K+infusion. The percent change in tidalvolume from a resting level is shown breath-by-breath immediately post K+ infusionin control animals (dotted line). In hyperosmotic animals, the percent change in tidalvolume from a resting level is shown first during 30 min of sucrose infusion, and thenthe ventilatory response to a K+ bolus is shown on a breath-by-breath basis.* significantly different from resting value.57DISCUSSIONPhysiological responses to acute changes in osmolalityAnalysis of acid-base status Calculated plasma bicarbonate apparently did notchange significantly during the hypertonic infusions, but since the pK of carbonic acid isaltered with ionic strength, these calculated bicarbonate values may be incorrect.Unfortunately, it is not possible to calculate with any accuracy the changes in pK duringthe hypertonic infusions with the available data. However, the pK should increase withNaC1 infusion and decrease with sucrose infusion, which would cause opposing changes incalculated bicarbonate in spite of very similar changes in pHa and PaCO2 during the twoprotocols. It was therefore felt that Stewart's approach to acid-base balance (Stewart,1983) could provide a useful analysis of the available data and insight into the cause of theextracellular acidosis. According to Stewart, in salt solutions such as blood andcerebrospinal fluid (CSF), [H+], [OW], and [HCO3 -] are dependent variables whoseconcentrations are determined by the three independent variables of the solution, whichare the strong ion difference ([SID]), total weak acid concentration, and PCO2. Despiteinherent technical and conceptual problems, this analysis of acid-base balance can offer aninformative approach to acid-base disturbances associated with large changes in electrolyteconcentrations. The only significant change in the measured independent variables in thisstudy in the first 45 min was a mean decrease of 5.8 meq•kg -1 in estimated [SID] duringhyperosmolality (Table I). This decrease was almost entirely due to excess CV in theextracellular fluid, as has been measured in mammals (Sotos et al., 1962, Makoff et al.,1970; Anderson and Jennings, 1988). The decrease in [SID] was independent of whetherplasma [Na] and [Cr] were increased (NaC1 infusion) or decreased (sucrose infusion).There is no doubt that this estimate of [SID] contains error, but the change in [SID] islarger than the possible error, and it does account for most of the measured pH change.Total weak acid concentration (protein) was not measured, but since plasma protein58concentration in ducks is very low, any change in protein concentration would be minorcompared to the change in [SID]. There was a significant increase in PaCO2 whichoccurred during all of the treatments. The rise in PaCO2 in the control animals wassmaller and occurred 45 min later than in the other animals, and did not measurably affectpHa. It is possible that the increase in PaCO2 was due to a decrease in parabronchialventilation during the relatively long restraint period.The cellular mechanism underlying the hyperchloremia during dilution acidosis isnot known. Hyperosmolality presumably causes ion transport alterations which decreasethe cation:anion ratio, resulting in acidemia. The acidosis is independent of whether thehyperosmolarity is produced with ionic or non-ionic substances. Based on the most recentstudies listed below, the term dilution acidosis is an inappropriate description of theacidosis caused by hyperosmolarity since neither dilution nor expansion of theextracellular space alone accounts for the observed acidosis. Hyposmolality with orwithout volume expansion has no measurable effect on pHa (Chang et al., 1975; seeRESULTS, Physiological responses to acute changes in osmolality), and volume contractionalkalosis occurring in association with chloride deficiency can be corrected by replacingCl- without restoring volume (Luke and Galla, 1983). Further studies of the ion transporteffects of hyperosmolality will be required in order to understand the nature of theacidosis.Constancy of plasma 1K+1 Plasma [K+] did not decrease significantly whenextracellular fluid volume was expanded with either hypertonic solution, indicating thattotal K+ increased slightly extracellularly (see Table II). The outward K+ shift issimultaneous with the decrease in pHa (Makoff et al., 1970; see RESULTS, Physiologicalresponses to acute changes in osmolality), and may be a passive readjustment to maintaina normal [Kli/[K le ratio. Other studies (Makoff et al., 1970; Wathen et al., 1982)have reported very large increases in extracellular [K+] with hypertonic saline andmannitol infusions, although the loads infused were significantly greater than in this study.5 9Such a K+ flux could change membrane potential in the direction of hyperpolarization,although the small magnitude of this efflux would probably render any membranepotential change physiologically insignificant. Two times normal tonicity leads to analkalinization of about 0.1 pH units (Whalley et al., 1991) and generally causes a smallhyperpolarization of 4-5 mV in muscle (Parker and Zhu, 1987; Whalley et al., 1991;Yamada, 1970). Hypertonicity does in fact reduce skeletal muscle contraction, although itis hypothesized to be the result of osmotic distortion of cell structure and nothyperpolarization (Parker and Zhu, 1987; Bruton, 1991). Such evidence suggests thatmembrane potential changes due to K+ flux are not significant during hyperosmotic stress.However, if this flux occurred in the carotid bodies, because the cells are small and have ahigh membrane resistance (Duchen et al., 1988), the opening of only a few K+ channelscould significantly affect membrane potential (Lynch and Barry, 1989). This remains tobe investigated.Respiratory compensation When evaluating the ventilatory response to an acidosis,the acid-base status of the extracellular fluid, intracellular fluid and the CSF (includingbrain interstitial fluid) must all be considered for their effect on chemoreceptor activity.In respiratory acidosis, arterial, CSF and intracellular pH all decrease (see Roos andBoron, 1981) and 'CTE increases significantly. During systemic metabolic acidosis, theresultant hyperventilation of spontaneously breathing animals can change CSF pHparadoxically because of the high permeability of the blood-CSF barrier to CO2 and itsrelatively low permeability to ions (Robin et al., 1958). Systemic pHe and pHi aredecreased, but CSF pH and brain pHi may not be, resulting in some attenuation of theventilatory response. During dilution acidosis, a concomitant, systemic, intracellularalkalosis may develop (see Fig. 7; Adler et al., 1975; Makoff et al., 1970). Potentialchanges to CSF pH and brain pHi are unknown, but there is clearly no accompanying,compensatory increase in ventilation. This is a novel observation, and while I am unable6 0to determine the reason for the lack of respiratory compensation from this first study, twopossibilities are worthy of discussion.The first possible explanation is that stimulation of respiration may be due solely tocentral stimulation of chemoreceptors, and there simply may have been a lack of CSF pHchange and, therefore, no central chemoreceptor stimulation during the dilution acidosis.However, in ducks, as in mammals, peripheral chemoreceptor contribution to acuteventilatory changes is approximately 20-40% (Milsom et al., 1981). Infusion of lacticacid in Pekin ducks, causing a systemic acidosis comparable to that seen in the presentstudy, immediately increased VE 225% (Jones and Shimizu, unpubl.). Thus, thesignificant systemic acidosis caused by hyperosmolality should also have increasedventilation, even if central chemoreceptors were not stimulated.A second possibility is that both peripheral and central chemoreceptors werestimulated by pHe changes, but there was concurrent development of a condition thatinhibited normal chemoreceptor stimulation. This possibility would be valid even if onlyperipheral receptors were affected. A significant difference between respiratory/metabolicacidosis and dilution acidosis in mammals, is the potential development of intracellularalkalosis during the latter perturbation. Evidence for a supression of ventilation byhyperosmolality is best supported by the response of three ducks to the infusion ofsucrose, as reported in the results of the present study. In the most extreme case, neithera decrease in pHa of 0.15 units nor a 16 torr increase in PaCO 2 was capable of stimulatingrespiration either peripherally or centrally under the hyperosmolal conditions, and 'TEactually decreased (Fig. 6). In ducks subjected to a period of submergence that resulted ina decrease in pHa of 0.2 pH units and an increase in PaCO2 from 32-57 Torr, VEincreased 350% upon emergence (Shimizu and Jones, 1987). In this case, both systemicand central pHe and pHi would be decreased. Manipulation of a potentially concurrentextracellular acidosis and intracellular alkalosis with respect to ventilation is done in thisstudy.61The effect of osmotic changes on chemoreceptor activity has been investigated onlybriefly. In cats, carotid body chemoreceptor activity increased when perfused in vivo withhyposmotic blood and decreased when perfused with blood made hyperosmotic withsucrose or NaC1 (Gallego and Belmonte, 1979). The minimal osmolality variationnecessary to obtain a detectable frequency change was 3-8% of the control. A 10%increase in osmolality decreased frequency about 20%, with generally greater reductionsobtained with sucrose. However, superfused carotid bodies in vitro responded in exactlythe opposite manner, with the authors concluding that the modifications in chemoreceptoractivity in vivo were produced by changes in carotid body blood flow due to a direct effectof hypo- and hyperosmotic solutions on vascular muscle tone. However, this does notexplain the in vitro results, unless presumably, the proposed cardiovascular effects in vivowere much greater than the direct effects on the receptors. In the present study, plasmaosmolality was increased 9-11 % and decreased 7% of resting values. Ventilatorydepression also tended to be slightly greater during hyperosmotic periods due to sucrosethan to NaC1 in this study. I did not observe any change in VE during hyposmolality, anda depression in VE in only a few animals during hyperosmolarity. It is possible that thestimulation of ventilation by the extracellular acidosis is approximately offset by adepression of ventilation due to an intracellular alkalosis during hyperosmolality, resultingin no ventilatory change. Gallego and Belmonte's protocol could easily have alloweddevelopment of an intracellular alkalosis without extracellular acidosis, since the carotidbodies were continuously flushed with fresh, hyperosmotic solution. Further work isrequired to fully explain these results. Increases in blood osmolality in ducks by 10mOsm (less than one-fifth of the osmotic change in this study) by either NaCl or mannitolalso caused a decrease in intrapulmonary chemoreceptor discharge frequencies at all levelsof PaCO2 (Adamson, 1984). The depressive effect became much larger at higher CO2levels. Fedde et al. (1982) have shown that intrapulmonary chemoreceptors in ducks aresensitive to venous blood changes, so consequently, these receptors could respond to6 2changes in osmolality of the blood. Hypertonic infusions have also been shown to inhibitthermal panting in mammals (Baker and Dawson, 1985), which again points to inhibitoryeffects of hyperosmolality on ventilation.Summary Analysis of the acid-base changes during hyperosmolality indicates thatwhile there is no significant change in calculated [HCO31, extracellular [C11 is greaterthan predicted. The significance of this with respect to development of the acidosis is asyet unknown. The lack of respiratory compensation to the acidosis suggests a suppressionof the predicted increase in chemoreceptor activity that normally occurs in response to anextracellular acidosis. It is hypothesized that an intracellular alkalosis develops at thesame time as the extracellular acidosis which may affect the ventilatory response.31P NMR of muscle and brainThe mean resting pectoral muscle pHi measured by 31p NMR (7.14+0.04; n=19)is in excellent agreement with the pHi measured previously in ducks in the same muscle(7.17; Stephenson and Jones, 1993), and within the range of mammalian skeletal musclepHi. The mean resting brain pHi (7.08+0.02) is also similar to values measured by 31PNMR in rats (Adler et al., 1990; Barrere et al., 1990). In calculating pHi, no attemptwas made to correct for any possible changes in intracellular ionic strength. Roberts et al.(1981) have pointed out that changes in intracellular ion concentrations within usualphysiological limits can significantly shift the Pi titration curve. Therefore, chemical shiftcould be altered by changes in intracellular conditions other than pH. Intracellular ionicstrength may have been altered in response to the protocol because of the shift ofintracellular water to the extracellular space and/or because of the selective addition ofmembrane permeable ions to the extracellular space. Osmolality and ion concentrationswere not measured in this study, but identical doses have been used in other experimentsin this thesis, and result in an increase in osmolality of 55 mOsm. Changes in pHi inresponse to hypertonicity measured with 14C-labeled 5,5-dimethyl-2,4-oxazolidinedione in63rat diaphragm muscle in vitro (Adler et al., 1975), and in erythrocytes of anesthetizeddogs in vivo (Makoff et al., 1970), were very similar during increases in osmolalitycomparable with those used in this experiment. The alkalinization in tissues in vitroappears to plateau at 0.1-0.15 pH units well before tonicity is doubled (Whalley et al.,1991). Therefore, it seems likely that any changes in ionic strength induced by theprotocol had a negligible effect on Pi chemical shift.Peripheral chemoreception The unique aspect of dilution acidosis is the lack ofrespiratory compensation. While ventilation was not directly measured in the NMRexperiments, the previous study in this thesis (see RESULTS, Physiological responses toacute changes in osmolality) showed that blood gases are reliable indicators of ventilatoryalterations under these experimental conditions. The changes in Pa0 2 and PaCO2measured in this study (see RESULTS, 31P nuclear magnetic resonance spectroscopy) arecomparable to those measured in the previous study (see RESULTS, Physiological responsesto acute changes in osmolality). Lactacidosis, typified by a decrease in both pHa and pHi,stimulated ventilation, whereas dilution acidosis, characterized by a comparable decreasein pHa but an increase in pHi, did not stimulate a ventilatory increase. The data suggestthat a decrease in pHi is required for an increase in ventilation in response to a change inpHa or PaCO2, therefore supporting a hypothesis that intracellular changes are necessaryfor chemoreception. The pHi changes measured here were not directly measured in thecarotid bodies, as mentioned previously, but since these organs are well perfused and, bynature of their function, sensitive to changes in blood chemistry, it seems reasonable tomake the assumption that the carotid body cells also responded to the osmotic changes andbecame alkalotic. In addition, the buffering capacity of muscle is very high while thebuffering capacity of carotid body tissue is very low (Wilding et al., 1992), meaning thatany pHi change would be even more pronounced in the carotid body than in muscle.As discussed previously (pp. 61-62), the few investigations that have beenpublished do support the suppression of chemoreceptor discharge and 'TE during6 4hyperosmolality. How hyperosmotic stress could affect chemoreceptor discharge is notclear. As many sensory receptors behave like mechanoreceptors, it is possible that thecarotid chemoreceptors could be responding simply to a distortion due to the movement ofwater. However, Lahiri (1977) has measured changes in carotid chemoreceptor dischargein vivo during intravenous hyperosmotic challenges, and concluded that because of the lagtime, they were not due to a mechanoreceptor response, but rather to some unidentified,intermediate step. Several studies have shown that increasing the osmolality of themedium bathing isolated muscle or nervous tissue causes intracellular alkalization,hyperpolarization and decreased electrical discharge (Adler et al., 1975; Abercrombie andRoos, 1983; Whalley et al., 1991), but, as previously discussed (pp. 58-59), thehyperpolarization is small and the significance is unclear. Depolarization and increaseddischarge during hyperosmolality have been measured in superfused, whole carotid bodyin vitro (Gallego et al., 1979; Gallego and Eyzaguirre, 1976). When the acid-basedisturbance is due to addition of an acid or base, intracellular alkalization in general isassociated with hyperpolarization, decreased neurotransmitter release and depressednervous discharge, including in carotid body glomus (Type 1) cells (Rigual et al., 1984;Eyzaguirre et al., 1989; He et al., 1991). There is at present no explanation for thedisparity in results with superfused in vitro carotid bodies. It is interesting to note thatduring recovery, pHi returned to resting levels while pHa was still significantly acidotic.However, ventilation as indicated by blood gases did not increase then in response to thearterial acidosis, suggesting that if internal alkalinization did inhibit chemotransduction, itdid so in a non-reversible or slowly reversible manner. Since hyperosmolality has somany known (and probably unknown) effects on cells, it is difficult to hypothesize whicheffect is responsible.Central chemoreception. Ventilation is controlled by both peripheral and centralchemoreceptors, and in ducks, as in mammals, central receptors control 60-80% of anacute ventilatory response to changes in PaCO2 (Milsom et al., 1981). There is still65considerable controversy over whether systemic, metabolic acid-base disturbances crossthe blood-brain barrier to any meaningful degree, although little work has been done onpH disturbances due to osmotic imbalance. Cserr et al. (1991) have shown that acute,systemic, hyperosmotic stress affects rat brain compartments selectively, decreasing brainextracellular volume and ion content, while the brain intracellular compartment maintainsits water content and gains electrolytes, indicating a degree of volume regulation. Theosmolality change and the time course of the experiment were comparable to that followedin this study. Although 3 of the 11 scans taken during the hyperosmotic infusion showeda significant alkalosis, there was no sustained significant increase in brain pHi found inthis study. The mean brain pHi during the sucrose infusion was 7.11+0.02, alkaloticcompared with the resting pHi. Although controversial, there may be, as Cserr's worksuggests, sufficient cell volume regulation to prevent any significant intracellular acid-basedisturbance (see Macknight et al., 1992), or, alternatively, pHi changes may be masked.The brain pHi monitored during this experiment was by necessity a global one since thesize of the coil resulted in sampling from most of the brain tissue. The brain isheterogeneous in cell type, and it is likely that cell pH is altered differently in differentbrain regions or cell groups.While PaCO2 usually increases about 4-5 torr during an hyperosmotic episode, ithas been noted that about 25% of animals subjected to hyperosmotic stress experienceextremely depressed ventilation and large pHa and blood gas changes, as observed in thestudy of the physiological responses to acute changes in osmolality in this thesis. Therewas also considerable inter-animal variation in PaCO2 in the brain study, with one animalhaving an increase in PaCO2 of 26.5 torr near the end of the infusion. Since CO2 isfreely permeable across the blood-brain barrier, this would normally stimulate centralreceptors to initiate a ventilatory increase. A ventilatory increase did not occur sincePa02 at this time was 27 torr lower than resting levels. If central receptors wereunaffected by the hyperosmolality (as the brain study suggests), and were responding to6 6this increase in CO2 , there must have been an opposing ventilatory depression arisingfrom some other source, presumably the peripheral receptors. Although I could notmeasure brain pH during lactic acid infusion, recent work has shown that a systemiclactacidosis of 0.36 pH units over 54 min did not alter brain pHi as measured by 31 P NMR(Adler et al., 1990), indicating that the primary ventilatory drive during lactacidosis inthis thesis was peripheral.Metabolism Normally, the intracellular compartment is well buffered againstextracellular pH variations, with pHi changes being only a fraction of the extracellularchange. In lactacidosis, the large drop in muscle pHi was transitory, with the acidosisthereafter being relatively small. This was not the case during hyperosmotic stress.Zeidler and Kim (1977) have provided evidence that during osmotic perturbation, theintracellular alkalinization causes structural instability of the cellular membrane band 3protein, which is associated with anion transport. Such a transport disturbance mightinhibit recovery in the short term.Perturbations in pHi of the magnitude measured in muscle in this study,particularly during hyperosmolality, could have a significant effect on metabolism (seeSomero, 1986). Although there were no changes in the relative concentrations of Pi, PCror ATP, proton concentration significantly decreased. Assuming that the creatine kinasereaction is in equilibrium, then the relative concentration of ADP presumably increased,and since ADP is considered to be a regulator of metabolism, this suggests that cellularrespiration may have increased during hyperosmolality. This conclusion is supported bythe increase in metabolism as measured by heat production reported in hyperosmoticallystressed muscle in aerobic conditions (Yamada, 1970). However, the metabolic responseto pHi perturbations are not always consistent (Chance and Conrad, 1959; Nioka et al.,1987).Summary It appears that brain pHi is little affected by acute systemichyperosmolality, but that the peripheral intracellular compartment undergoes a significant67alkalosis during extracellular dilution acidosis. Exogenously produced lactacidosis resultsin a decrease in systemic pHi. The data suggest that intracellular pH may play a role ininitiating ventilatory changes since normal respiratory compensation to the extracellularacidosis is prevented during hyperosmolality but not during lactacidosis. It is suggestedthat there is a ventilatory depression generated peripherally during acute systemichyperosmotic stress.31P NMR of pectoral muscle during ion-exchange blockadeDosage DIDS is a stilbene disulfonate that binds specifically and irreversibly tocell membrane band 3 protein to inhibit anion transport in a variety of cell types.Assuming that the distribution of DIDS is primarily extracellular and that the extracellularspace of ducks is about 25% body weight (Ruch and Hughes, 1975), then the DIDSconcentration in the extracellular fluid would have been about 1.2 mmo1•1 -1 , or 1000 timesthe Ki (concentration resulting in 50% inhibition of ion transport) for DIDS inerythrocytes (1.2 umo1.1 -1). This concentration is similar to that used in some otherstudies (Deng and Johanson, 1989; Javaheri et al., 1984), but is generally higher than thatused in most CSF studies (Nattie and Adams, 1988; Nishimura et al., 1988). DIDS isirreversible and would have remained effective over the 85 min protocol.The dosage of amiloride is similar to that used by other investigators (Altenberg etal., 1989; Obika, 1989). Using the same assumptions about extracellular space anddistribution described above, then the amiloride concentration in the extracellular fluidwould have been about 0.15 mmo1•1 -1 . Amiloride is an effective inhibitor of the Na+/H+antiporter at a concentration of 0.1-1 mM when extracellular [Nat] is in the physiologicalrange (Fitzgerald et al., 1990). The action of amiloride is immediate but reversible, andit is possible that the slight diuresis caused by sucrose infusion may have resulted in aneffective reduction in the amiloride concentration. While this may have occurred, itwould be somewhat offset by the improved effectiveness of amiloride due to the reduction68in extracellular [Nal (Mahnensmith and Aronson, 1985) caused by the sucrose infusion.Since pHi decreased due to amiloride and did not recover during the experiment, thissuggests that the amiloride was effectively present throughout the study.Control of ventilation Infusion of either DIDS or amiloride caused a decrease inboth pHa and pHi, and an increase in 'E. It has previously been reported that theventilatory response to CO2 after intracerebroventricular DIDS infusion is significantlyenhanced (Adams and Johnson, 1988; Nattie and Adams, 1988). Extracellular pHregulation (and therefore presumably intracellular as well) is considered to be lesseffective after DIDS treatment (Nattie and Adams, 1988; Ahmad and Loeschcke, 1983),although no study has directly shown that the enhanced ventilatory responsiveness is dueto the altered pH regulation. The results of this study strongly support that latterconclusion. The change in pHi, which normally becomes alkalotic during extracellularhyperosmolality, was reversed after DIDS treatment, while VE, which normally does notincrease in response to the extracellular acidosis during hyperosmolality, increasedsignificantly after DIDS treatment. Essentially the same is true for amiloride treatmentunder these conditions, although Fitzgerald et al. (1990) have shown that amiloride has noeffect on the response of chemoreceptor discharge to hypercapnia (but significantlydepressed the response to hypoxia). This may indicate that C1 -/HCO3 - exchange is ofprimary importance in pH regulation during hypercarbia.The contribution of pHi to the ventilatory response is illustrated in Fig. 18. Theslope of the ventilatory response during hyperosmolality is clearly different from theresponse when pHa and pHi are coupled. The results of this study however, suggest that adecrease in pHi is not essential for chemoreceptor stimulation, since VE significantlyincreased in DIDS-treated animals when pHa was acidotic but pHi had not yet changed(Figs. 9 and 10). In amiloride-treated animals, pHi began to recover while pHa continuedto decrease, yet VE continued to increase. It is also interesting that drugs that inhibit theNa+-dependent H+-extruding systems involved in cell pH recovery, also inhibit dopamine•69350 -300 -250 -200 r =0.5C150E100 ->50 -0--50-0.15^-0.10^-0.05^0.00pHe350 -r=0.8 O.••0•017)^200 -^0^ •150 -^06^100 -^r = 0.4•>• .......... •••^r-0.60--50-0.2^-0.1^0.0^0.1^0.2pHiFig. 18. The relationship between ventilation and intra- and extracellular pHduring hyperosmolality. Filled circles represent DIDS-treated animals; opencircles represent amiloride-treated animals; diamonds represent hyperosmolalityalone. Lines are regressions through the points.0 •• • +.^r = 0.4•300 -250 -O •50 -7 0release (Rocher et al., 1991), which should suppress nervous discharge and ventilatoryincreases. Since there was a substantial ventilatory increase with sucrose infusion in theamiloride-treated animals in this study, dopamine release might posibly have beenaugmented by hyperosmolality under these conditions despite the recovering pHi.Therefore, a drop in extracellular pH alone is apparently sufficient to stimulatechemoreceptors to increase ventilation. However, chemoreceptors do appear to respond topHi, since an intracellular alkalosis under otherwise similar conditions suppresses theventilatory increase. A role of pHi in control of chemoreceptor function has beenproposed before (Hanson et al., 1981; Adler et al., 1975; Lassen, 1990), but to date therehas been no evidence to either support or negate this hypothesis. There was an increase inVE with DIDS infusion alone which occurred without any significant change in pHa orpHi in this study (Figs. 9 and 10). This has also been observed by Adams and Johnson(1988) and is somewhat confusing. Presumably this response is mediated by thechemoreceptors, but the mechanism is unknown. The concentration of DIDS alone in thisprotocol did not disrupt pH regulation enough under resting control conditions to cause ameasurable change in pH, but was sufficient to alter regulation during an acid-base/respiratory challenge.It is interesting to note the changes in blood gases during both treatments andduring hyperosmolality alone. PaC O2 always increased significantly duringhyperosmolality by about 4 torr (Figs. 5 and 15). This could occur if reverse C1 -/HCO3 -exchange (Cl - efflux, HCO3 - influx) was augmented, which would inhibit CO 2 hydrationintracellularly, thus increasing PaCO2, or it may be due to the increased metabolic rateduring hyperosmolality. Amiloride-treated animals showed a significant drop in PaCO2consistent with increased ventilation. PaCO 2 did not decrease in DIDS-treated animalsdespite the large increase in ventilation because the inhibition of CF/HCO3 - exchangewould inhibit CO 2 hydration.7 1Ion exchange in single cells During in vivo hyperosmotic stress, calculationsshowed that extracellular [Na+] is not decreased beyond what is accounted for by dilutiondue to the sucrose infusion and intracellular water flux, while extracellular [Cl -] wassignificantly higher than expected (Makoff et al., 1970; Anderson and Jennings, 1988; seeRESULTS, Physiological responses to acute changes in osmolality). This resulted in adecrease in [SID] and suggested the stimulation of the c1-aico3- exchanger followinghyperosmotic shrinkage. However, much of the in vitro work done on osmotic activationof ion exchange implicates the stimulation of Na+/H+ exchange by hyperosmotic stress asthe primary ion regulating mechanism through increased intracellular proton affinity anddecreased intracellular Na+ affinity (Green et al., 1988; Green and Muallem, 1989;Whalley et al., 1991). Furthermore, most studies in volume regulating cells indicate thatosmotic shrinkage generally leads to net Na+ and CV uptake (see Hoffmann andSimonsen, 1989), the latter being completely blockable by DIDS or SITS (4-acetamido-4'-isothiocynao-2,2'-stilbene disulfonic acid). In contrast, Abercrombie & Roos (1983)measured significantly lower Cl- activity in hypertonic frog muscle, implying a netefflux. Zeidler and Kim (1977) have also shown that calf and cow red cells exposed tohypertonic media undergo a net loss of CF accompanied by an intracellular alkalinizationwhich was completely inhibited by SITS. DIDS was very effective in preventingintracellular alkalinization due to sucrose infusion in this study, although there was aprominent time lag, and it also equalized the dilution of Na+ and Cl- (Table II).However, the decrease in extracellular [C11 was even less than during the same osmoticload alone, the opposite of what would be predicted. This suggests that CF was stillcoming out of the cell by the remaining functional Cl/HCO3 - exchange plus some othertransport mechanism(s). There was also some Na+ efflux (extracellular [Nat] is about 8mEq higher than predicted), but no K+ efflux. The extracellular dilution of K+ suggeststhat the small intracellular hyperpolarization that normally occurs during hyperosmolalitydue to K+ efflux would not have developed in the hyperosmotic, DIDS-treated animals.7 2This could be a factor in the ventilatory increase observed in the latter animals, as well asin amiloride-treated animals.The results of the pH regulation in response to hyperosmolality are made moredifficult to interpret because of the significant fall in pHi due to amiloride (also observedby Whalley et al., 1991). However, pHi remained low in response to hyperosmotic stressafter amiloride, suggesting a role for Na+/H+ exchange in the alkalinization processnormally accompanying cell shrinkage. It is curious, however, that extracellular [Nat] isnot higher than in animals treated only with sucrose (Table II). In addition, because pHireturned to values not significantly different from resting levels after 21 min of sucroseinfusion, it is apparent that one or more additional exchange mechanisms are alsoinvolved. Amiloride also had the surprising effect of reducing Cl- efflux (or increasingCl- influx), so that the Na+ and Cl- were equally diluted.While both blockers inhibited contraction alkalosis, the ion exchange mechanismsduring hyperosmotic stress are extremely complicated and are difficult to predict in thewhole animal or to compare to other studies since ion exchangers are both cell and speciesspecific. It appears that Na+, Cl - and K+ all play a prominent role in the hypertonicresponse, but the mechanisms are too complicated to identify at this whole animal level.Further single cell experiments monitoring all the major ions are required.Total CO2 The significant decrease in total CO2, combined with the significantincrease in PaCO2 that always accompanies acute hyperosmolality, indicates that theHCO3 - concentration is decreasing during the osmotic challenge. This confirms withoutdoubt that the acid-base disturbance associated with acute hyperosmolality is not simply arespiratory acidosis, and again implies that C1 -/HCO3- exchange may play a primary rolein this perturbation.Summary Since both DIDS and amiloride had the same effect on the normalresponse to hyperosmolality (both drugs reversed the intracellular alkalosis to anintracellular acidosis, and did not prevent the extracellular acidosis), it indicates that both7 3Table II. Decreases in electrolytes (meq•kg- ') during acute hyperosmolality alone andhyperosmolality after various treatments. Data are shown as the changes from restinglevels after approximately 30 min of sucrose infusion. Amount of sucrose infused per kganimal is identical in each study except the first, where the load infused was smaller. Seetext for details.A C1- A Na+ A K÷Sucrose 11.4±2.1 16.1±2.0 0.13±0.16Sucrose after amiloride 20.9+2.9 20.2+12.2 0.65 +0.32Sucrose after DIDS 9.9+1.6 11.7±4.6 0.64+0.39Sucrose (after hypercapnia) 13.6+1.6 21.4±1.1 0.20+0.17Sucrose (after hypoxia) 12.9+0.8 20.9+1.1 0.16+0.377 4ion exchange mechanisms are crucial to the complete perturbatory effects ofhyperosmolality. However, the unusual changes in the measured extracellular ionconcentrations suggests that the ion exchange disruptions are complex. The data stronglysuggest that intracellular pH plays a role in initiating ventilatory changes since normalrespiratory compensation to the extracellular acidosis is prevented when the intracellularmilieu is alkalotic but initiated when the intracellular milieu is acidotic or homeostatic.Responses to respiratory stimuliAs the previous studies have shown, the extracellular acidosis induced byhyperosmolality did not stimulate an increase in ventilation. This phenomenon has led tothe hypothesis that the intracellular contraction alkalosis that develops systemicallyconcurrent to the dilution acidosis supresses the normal chemoreceptor response to theextracellular acidosis, thus resulting in no ventilatory change. A decrease in pHa ofsimilar magnitude and time course is sufficient to double ventilation when the acidosis iscaused by lactic acid infusion, which also decreases intracellular pH. While it has notbeen fully resolved how systemic changes in osmolality affect the various fluidcompartments of the brain (cerebrospinal, interstitial, and intracellular fluids), theavailable data suggests that there are minimal effects on the brain intracellularcompartment (Cserr et al., 1991; this thesis). This suggests that the primary effects ofsystemic hyperosmolality on ventilation take place via the peripheral chemoreceptors. Theinterpretation of the ventilatory data in this study must take into account that the responseto hypercapnia is both periperally and centrally mediated, while the response to hypoxiaand K+ is essentially solely under peripheral control.Hypercapnia The CO2 response curve in Pekin ducks is linear over the range ofinterest in this study (30-50 torr) (Dodd and Milsom, 1987; Bouverot et al., 1974), so a1st-order regression line was fitted to the data in Fig. 17. The absolute VE values in the75present study are somewhat lower than in the above studies, but the slopes of the control(normosmotic) CO2 response curves are similar.Acute metabolic acidosis usually causes a displacement of the CO 2 response curveto significantly higher ventilatory levels (Saito et al., 1960; Schuitmaker et al., 1986).This does not occur during dilution acidosis. Hyperosmolality clearly caused a shift of theCO2 ventilatory response curve to the right, implying an increase in the ventilatorythreshold to CO2 (ventilatory threshold being the level of PaCO2 at which here is asignificant increase in VE from rest), and also slightly decreased the sensitivity to CO 2(Fig. 16). Such a displacement of the CO 2 curve explains why there is no ventilatoryresponse to the approximately 4 torr increase in PaCO2 generated during the hyperosmoticinfusion itself.Compared with normosmotic hypercapnic animals, depressed chemoreceptordischarge has been observed in intrapulmonary chemoreceptors in ducks during saline ormannitol-induced hyperosmolality at all levels of PaCO2, with the effect increasing as thelevel of inspired CO2 was raised (Adamson, 1984). While these changes inintrapulmonary chemoreceptor discharge would have little effect on TE, being primarilyinvolved in determining respiratory pattern (Milsom et al., 1981), this effect could also beoccurring in the carotid bodies, an interpretation which would support the ventilatorychanges observed in this study. The 3.5% inspired CO 2 in this study would inhibitintrapulmonary chemoreceptor discharge only about 25% (Fedde et al., 1974), thereforethe VE changes seen are probably not due simply to an inabilty to maintain an adequaterespiratory pattern. Depressed ventilatory responses to inspired CO2 have also beenobserved in ducks after injection with acetazolamide (Andersen and Hustvedt, 1967;Powell et al., 1978), although it is not clear how hyperosmolality could affect carbonicanhydrase.Acute hyperosmolality has not been shown to have any consistent, significant effecton the brain intracellular compartment (Cserr et al., 1991; this thesis). The 2 min7 6hypercapnic challenge administered in this study should have stimulated centralchemoreceptors (Jones and Purves, 1970) to the same magnitude both before and duringhyperosmolality. While much of the ventilatory response to CO2 is centrally mediated(60-80%), it has been suggested that the arterial chemoreflex drive is essential for normalventilatory responses to CO2 since carotid body denervation shifts the ventilatory responsecurve to the right and decreases the sensitivity to CO2 (Bouverot et al., 1974; Lahiri etal., 1978). The results of this study are consistent with this conclusion, although it cannot account for the lack of ventilatory increase to the extreme elevations of PaCO2 seen insome ducks (see RESULTS, Physiological responses to acute changes in osmolality, and 31Pnuclear magnetic resonance spectroscopy (Brain)). These cases suggest that there is alsosome depression of the peripheral drive during hyperosmolality, particularly if the centralchemoreceptor threshold for CO 2 is normally lower than that of the peripheralchemoreceptors (Lahiri et al., 1978). Potentially, the systemic intracellular alkalosis thatdevelops during acute hyperosmolality may suppress chemoreceptor discharge. If centralchemoreceptors respond to CO 2 via changes in [11±], as is accepted both generally (seeMilsom, 1990 for review) and specifically in peripheral receptors (see Gonzalez et al.,1992 for review), then the trend toward an intracellular alkalosis centrally could alsodecrease the ventilatory threshold to CO2, without affecting the sensitivity.Hypoxia and IC+ There was no apparent change in sensitivity to hypoxia aftersucrose infusion, but the 02 response curve was shifted upward compared with thecontrol, normosmotic response (Fig. 15). However, the 02 response curve for ducks isactually an exponential-shaped curve (Jones and Holeton, 1972; Bouverot and Sebert,1979), rather than a straight line as shown in Fig. 15. Since the data generated in thisexperiment were insufficient to draw a curve, there is a possibility that the points between40-45 torr are merely on the steep region of the curve and that there is no upward shift.This is unlikely for the following four reasons. First, the 02 response curve in ducks isnot so steep that, on average, a 2.8+6.1 torr difference in Pa02 would account for VE77increasing almost 200% over resting compared with a 100% increase in VE in controlanimals (see Jones and Purves, 1970; Jones and Holeton, 1972). In addition, thenormosmotic and the hyperosmotic Pa02 values after 2 min of hypoxia are notsignificantly different. Second, Jones and Holeton (1972) noted that the increase in VE incontrol animals terminated around a Pa02 of 47 ton, at which point both in their studyand in this study (where Pa02 was actually 44.6+2.3 ton), VE was approximately doubleresting YE. The same hypercapnic challenge in hyperosmotic animals led to a similardrop in Pa02 (41.8+2.1 torr) but resulted in almost a tripling of VE in this study. Incontrast, a further decrease in Pa02 in Jones and Holeton's ducks to 38 torr actuallydecreased YE. It seems clear that the ventilatory response to hypoxia was significantlyaffected by hyperosmolality. This conclusion is further supported by the fact that carotidbody discharge increases in hyperosmotic cats during hypoxia (Lahiri, 1977). Finally, theresponse to K+ confirms the hypoxic data, particularly since K+ may be associated withthe hypoxic response (Lopez-Lopez et al., 1989; Ganfornina and Lopez-Barneo, 1991)and, like hypoxia, is apparently not dependent upon a pH change. The response to the K+bolus demonstrated that hyperosmolality was significantly increasing the ventilatoryresponse to these stimuli (Fig. 17). It is not surprising that there was no ventilatoryresponse to the slow K+ infusion, since there was no measurable increase in plasma [K+](the excess ion in the blood is quickly absorbed by muscle and bone), and a fast rate ofarterial [K+] increase is crucial to initiate a ventilatory response. Increased ventilatoryresponse to hypoxia with increased [H+] has long been established both in vivo (Natsui,1970; Gabel and Weiskopf, 1975), and in carotid bodies in vitro (Biscoe and Duchen,1990b), while chemoreceptor discharge in response to hypoxia during alkalosis isabolished (Eyzaguirre and Koyano, 1965). It is unclear, however, whether intra- orextracellular pH is critical since intracellular pH is rarely measured. Since it is knownthat there is a systemic intracellular alkalosis during dilution acidosis, the present resultssuggest that there is some interaction between the extracellular pH and hypoxia at the level78of the carotid bodies. As chemoreceptive mechanisms have not yet been conclusivelyelucidated, it is very difficult to speculate on the nature of any interaction.Metabolism There was a small but significant increase in metabolic rate duringhyperosmolality during both ventilatory challenges (which supports the 31P NMR datadiscussed above, pg. 66). Whole animal metabolic rate is known to increase duringextracellular alkalosis and decrease during extracellular acidosis. It has been demonstratedthat changes in V02 are dependent upon the extracellular pH and are inversely related toPaCO2 (see Patterson and Sullivan, 1978). However, Patterson and Sullivan (1978) pointout that an intracellular site of action is likely and has, in fact, not been disproved. In thecurrent study, the opposing changes in pH during hyperosmolality and the rise in V02clearly support an intracellular pH influence. Alkalosis stimulates glycolysis, primarilythrough effects on the activity of the regulatory enzyme phosphofructokinase (Fidelman etal., 1982), or the increased metabolic rate could reflect an increase in ion transport. Theelevated metabolic rate may also have contributed to the increase in PaCO2 measuredduring the hyperosmotic challenge, especially since there was no modulationg influence ofincreased ventilation.Summary Hyperosmolality decreased the normal ventilatory response tohypercapnia, but increased the normal respiratory response to hypoxia and K+. The shiftsin the two gas response curves in opposite directions have several interesting implications.First, it indicates that the mechanisms of chemoreception for CO 2 and 02 are different, aconcept which is generally supported in the literature but which has not been conclusivelydemonstrated (see Gonzalez et al., 1992). Furthermore, stimulation of thechemoreceptors by K+ and low 02 may have some basic mechanism in common, and itsupports the involvement of K+ in the hypoxic response (Lopez-Lopez et al., 1989).Finally, the changes in the ventilatory responses to the stimuli, combined with theopposite intra- and extracellular pH changes, imply that both intra- and extracellular pHhave a role in controlling ventilation.7 9ConclusionsThe data indicate that intracellular pH may play a role in initiating ventilatorychanges since normal respiratory compensation to extracellular acidosis is preventedduring hyperosmolality but not during lactacidosis. This is a unique observation, and it isconcluded that there is a ventilatory depression generated peripherally during acutesystemic hyperosmotic stress. Since hyperosmolality can have a number of effects oncells, chemoreceptor suppression could be due to intracellular alkalosis, hyperpolarization,cell structure disruption, decreased dopamine release or other perturbations not yetunderstood. However, the acid-base disturbance associated with acute hyperosmolalityappears to offer a unique system to further investigate the probability that chemoreceptorsrespond to changes in both intra- and extracellular pH.8 0REFERENCESAbercrombie, R.F. and Roos, A. (1983). The intracellular pH of frog skeletal muscle: itsregulation in hypertonic solutions. J. Physiol. (Lond.) 345: 189-204.Acker, H. and C. Eyzaguirre (1987).Light absorbance changes in the mouse carotid bodyduring hypoxia and cyanide poisoning. Brain Res. 409: 380-385.Acker, H., E. Dufau, J. Huber and D. Sylvester (1989). Indications to an NADPHoxidase as a possible p02 sensor in the rat carotid body. FEBS Lett. 256: 75-78.Adams, J.M., and Johnson, N.L. (1988). Intracisternal DIDS enhances ventilatoryresponse to rebreathing CO2 in rats. J. Appl. Physiol. 65: 1611-1616.Adamson, T.P. (1984). Discharge characteristics of intrapulmonary chemoreceptors in thePekin duck. Ph.D. thesis, Univ. of California Davis. 92 pp.Adler, A., Anderson, B. and Zett, B. (1975). Effect of osmolarity on intracellular pH ofrat diaphragm muscle. Am. J. Physiol. 228: 725-729.Adler, S., V. Simplaceanu and C. Ho (1990). Brain pH in acute isocapnic metabolicacidosis and hypoxia: a 31P-nuclear magnetic resonance study. Am. J. Physiol. 258: F34-F40.Ahmad, H.R. and H.H. Loeschcke (1983). Evidence for a carrier mediated exchangediffusion of HCO3 - against Cl- at the interphases of the central nervous system. In:Central Neurone Environment, ed. by M.E. Schlafke, H.P. Kepchen and W.R. See.Springer-Verlag, Berlin. pp. 13-21.Altenberg, G.A., P.C. Aristimuno, C.E. Amorena and A.C. Taquini (1989). Amilorideprevents the metabolic acidosis of a KC1 load in nephrectomized rats. Clin. Sci. 76: 649-652.Anichkov, S.V. and M.L. Belenkii (1963). Pharmacology of the Carotid BodyChemoreceptors. Pergamon Press.Andersen, H.T. and B.-E. Hustvedt (1967). Carbon dioxide excretion and pH-variationsin diving ducks after carbonic anhydrase inhibition. Acta Physiol. Scand. 69: 203-208.Anderson, J.W. and Jennings, D.B. (1988). Osmolality, NaCl dietary intake, andregulation of ventilation by CO2. Am. J. Physiol. 255, R105-R112.81Asano, S., E. Kato, M. Yamauchi, Y. Ozawa, M. Iwasa, T. Wada and H. Hasegawa(1966). The mechanism of the acidosis caused by infusion of saline solution. Lancet 1:1245-1246.Baker, M.A. and D.D. Dawson (1985). Inhibition of thermal panting by intracarotidinfusion of hypertonic saline in dogs. Am. J. Physiol. 249: R787-R791.Barrere, B., P. Meric, J. Borredon, G. Berenger, J.-C. Beloeil and J. Seylaz (1990).Cerebral intracellular pH regulation during hypercapnia in unanesthetized rats: a 31 Pnuclear magnetic resonance spectroscopy study. Brain Res. 516: 215-221.Biscoe, T.J. and M.R. Duchen (1990a). Cellular basis of transduction in carotidchemoreceptors. Am. J. Physiol. 258:Biscoe, T.J. and M.R. Duchen (1990b). Responses of type I cells dissociated from therabbit carotid body to hypoxia. J. Physiol. (Lond.) 428: 39-59.Bruton, J.K. (1991). Contraction of rat coleus muscle and the effects of hypertonictreatments. Jap. J. Physiol. 41: 189-201.Bouverot, P. and Ph. Sebert (1979). 0 2-chemoreflex drive of ventilation in awake birds atrest. Respir. Physiol. 37: 201-218.Bouverot, P., N. Hill and Y. Jammes (1974). Ventilatory responses to CO2 in intact andchronically chemodenervated Peking ducks. Respir. Physiol. 22: 137-156.Chance, B. and H. Conrad (1959). Acid-linked functions of intermediates in oxidativephosphorylation. II. Experimental studies of the effect of pH upon respiratory,phosphorylative, and transfer activities of liver and heart mitochondria. J. Biol. Chem.234: 1568-1570.Chang, B.S., S.I. Kahn and S. Garella (1975). Influence of tonicity on bicarbonategeneration in volume expansion. Clin. Res. 23: 357A.Cserr, H.F., M. DePasquale, C. Nicholson, C.S. Patlak, K.D. Pettigrew and M.E. Rice(1991). Extracellular volume decreases while cell volume is maintained by ion uptake inrat brain during acute hypernatremia. J. Physiol. (Lond.) 442: 277-295.De Castro, F. and M. Rubio (1968). The anatomy and innervation of the blood vessels ofthe carotid body and the role of chemoreceptive reactions in the autoregulation of theblood flow. In: Arterial Chemoreceptors, ed. by R.W. Torrance. Blackwell Press,Oxford. pp. 267-277.8 2Deng, Q.-S. and Johanson, C.E. (1989). Stilbenes inhibit exchange of chloride betweenblood, choroid plexus and cerebrospinal fluid. Brain Res. 501, 183-187.Dodd, G.A.A. and W.K. Milsom (1987). Effects of H+ versus CO2 on ventilation in thePekin duck. Respir. Physiol. 68: 189-201.Duchen, M.R., K.W.T. Caddy, G.C. Kirby, D.L. Patterson, J. Ponte and T.J. Biscoe(1988). Biophysical studies of the cellular elements of the rabbit carotid body. Neurosci.26: 291-311.Eyzaguirre, C. and H. Koyano (1965). Effects of hypoxia, hypercapnia, and pH on thechemoreceptor activity of the carotid body in vitro. J. Physiol. (Lond.) 178: 385-409.Eyzaguirre, C., L. Monti-Bloch, M. Baron, Y. Hayashida, and J.W. Woodbury (1989).Changes in glomus cell membrane properties in response to stimulants and depressants ofcarotid nerve discharge. Brain Res. 477: 265-279.Fedde, M.R., R.N. Gatz, H. Salama and P. Scheid (1974). Intrapulmonary CO2 receptorsin the duck: I. Stimulus specificity. Respir. Physiol. 22: 99-114.Fedde, M.R., J.P. Kiley, F.L. Powell, and P. Scheid (1982). Intrapulmonary CO 2receptors and control of breathing in ducks: effects of prolonged circulation time tocarotid bodies and brain. Respir. Physiol. 47: 121-140.Fidelman, M.L., S.H. Seeholzer, K.B. Walsh and R.D. Moore (1982). Intracellular pHmediates action of insulin on glycolysis in frog skeletal muscle. Am. J. Physiol. 242: C87-C93.Fitzgerald, R.S., M. Shirahata and S. Lahiri (1990). Amiloride and carotid bodychemoreception of hypercapnia and hypoxia. Respir. Physiol. 81, 337-348.Gabel, R.A. and R.B. Weiskopf (1975). Ventilatory interaction between hypoxia and[H+] at chemoreceptors of man. J. Appl. Physiol. 39: 292-296.Galla, J.H. and R.G. Luke (1988). Chloride transport and disorders of acid-base balance.Ann. Rev. Physiol. 50: 141-158.Galla, J.H., D.N. Bonduris, P.W. Sanders and R.G. Luke (1984). Volume-independentreductions in glomerular filtration rate in acute chloride-depletion alkalosis in the rat. J.Clin. Invest. 74: 2002-2008.Gallego, R. and C. Belmonte (1979). The effects of blood osmolality changes on catcarotid body chemoreceptors in vivo. Pfleugers Arch. 380: 53-58.8 3Gallego, R. and C. Eyzaguirre (1976). Effects of osmotic pressure changes on the carotidbody of the cat in vitro. Fed. Proc. 35: 404.Gallego, R., C. Eyzaguirre, and L. Monti-Bloch (1979). Thermal and osmotic responsesof arterial receptors. J. Neurophysiol. 3: 665-680.Ganfornina, M.D. and J. Lopez-Barneo (1991). Single K+ channels in membrane patchesof arterial chemoreceptor cells are modulated by 0 2 tension. Proc. Natl. Acad. Sci. USA88: 2927-2930.Garella, S., B.S. Chang and S.I. Kahn (1975). Dilution acidosis and contraction alkalosis:review of a concept. Kidney Mt. 8: 279-283.Green, J. and S. Muallem (1989). A common mechanism for activation of the Na+/H+exchanger by different types of stimuli. FASEB J. 3: 2408-2414.Green, J., D.T. Yamaguchi, C.R. Kleeman and S. Muallem (1988). Selectivemodification of the kinetic properties of Na+/H+ exchanger by cell shrinkage andswelling. J. Biol. Chem. 263: 5012-5015.Gonzalez, C., L. Almaraz, A. Obeso and R. Rigual (1992). Oxygen and acidchemoreception in the carotid body chemoreceptors. Trends Neurosci. 15: 146-153.Hanson, M.A., P.C.G. Nye and R.W. Torrance (1981). The exodus of an extracellularbicarbonate theory of chemoreception and the genesis of an intercellular one. In: ArterialChemoreceptors, Proceedings of the VI International Meeting, ed. by C. Belmonte, D.J.Pallot, H. Acker and S. Fidone. Leicester University Press, Oxford. pp. 403-416.He, A.-F., J.-Y. Wei, and C. Eyzaguirre (1991). Intracellular pH and some membranecharacteristics of cultured carotid body glomus cells. Brain Res. 57: 258-266.Helbecka, N.V.L., J.L. Casterline, C.J. Smith and C.S. Shaffner (1964). Investigation ofplasma carbonic acid pK of the chicken. Poultry Sci. 43: 138-144.Hoffman, E.K. and L.O. Simonsen (1989). Membrane mechanisms in volume and pHregulation in vertebrate cells. Physiol. Rev. 69: 315-382.Javaheri, S., J. Weyne, G. Demeester and I. Leusen (1984). Effects of SITS, an aniontransport blocker, on CSF ionic composition in metabolic alkalosis. J. Appl. Physiol. 57:92-97.8 4Jean, T., C. Frelin, P. Vigne and M. Lazdunski (1986). The Na+/H+ exchange system inglial cell lines. Properties and activation by hyperosmotic shock. Eur. J. Biochem. 160:211-219.Jones, D.R. and G.F. Holeton (1972). Cardiovascular and respiratory responses of ducksto progressive hypocapnic hypoxia. J. exp. Biol. 56: 657-666.Jones, D.R. and M.J. Purves (1970a). The carotid body in the duck and the consequencesof its denervation upon the cardiac responses to immersion. J. Physiol. (Lond.) 211: 295-309.Jones, D.R. and M.J. Purves (1970b). The effect of carotid body denervation upon therespiratory response to hypoxia and hypercapnia in the duck. J. Physiol. (Lond.) 211:295-309.Kost, G.J. (1990). pH standardization for phosphorus-31 magnetic resonance heartspectroscopy at different temperatures. Mag. Res. Med. 14: 496-506.Lahiri, S. (1977). Introductory remarks: oxygen linked response of carotidchemoreceptors. Adv. Exp. Med. Biol. 78: 185-202.Lahiri, S., A. Mokashi, R.G. Delaney and A.P. Fishman (1978). Arterial PO2 and PCO2stimulus threshold for carotid chemoreceptors and breathing. Respir. Physiol. 34: 359-375.Lassen, N.A. (1990). Is central chemoreceptor sensitive to intracellular rather thanextracellular pH? Clin. Physiol. 10: 311-319.Lopez-Lopez, J., C. Gonzalez, J. Urena and J. Lopez-Barneo (1989). Low P02selectively inhibits K channel activity in chemoreceptor cells of the mammalian carotidbody. J. Gen. Physiol. 93: 1001-1015.Luke, R.G.and J.H. Galla (1983). Chloride-depletion alkalosis with a normal extracellularfluid volume. Am. J. Physiol. 245: F419-F424.Lynch, J.W. and P.H. Barry (1989). Action potentials initiated by single channels openingin a small neuron (rat olfactory receptor). Biophys. J. 55: 755-768.Macknight, A.D.C., J. Grantham and A. Leaf (1992). Physiologic and pathophysiologicresponses to changes in extracellular osmolality. In: The Kidney: Physiology andPathophysiology, 2nd ed., ed. by D.W. Seldin and G. Giebisch. Raven, New York. pp.1779-1806.8 5Mahnensmith, R.L. and P.S. Aronson (1985). The plasma membrane sodium-hydrogenexchanger and its role in physiological and pathophysiological processes. Circ. Res. 56:773-788.Makoff, S.L., A. Da Silva, B.J. Rosenbaum, S.E. Levy, and M.H. Maxwell (1970).Hypertonic expansion: acid-base and electrolyte changes. Am. J. Physiol. 218: 1201-1207.Maskrey, M. and D. Trenchard (1989). A vagally mediated response to metabolic acidosisin anesthetized rabbits. J. Appl. Physiol. 66: 1635-1641.McCormack, J.G., A.P. Halestrap and R.M. Denton (1990). Role of calcium ions inregulation of mammalian intramitochondrial metabolism. Physiol. Rev. 70: 391-425.Milsom, W.K. (1990). Control and co-ordination of gas exchange in air breathers. In:Advances in Comparative and Environmental Physiology, Vol. 6., ed. by R.G. Boutilier.Springer-Verlag, Berlin. pp. 347-400.Milsom, W.K., D.R. Jones and G.R.J. Gabbot (1981). On chemoreceptor control ofventilatoy responses to CO2 in unanesthetized ducks. J. Appl. Physiol. 50: 1121-1128.Natsui, T. (1970). Respiratory response to arterial H+ at different levels of arterial PCO2during hyperoxia or hypoxia. Pfleugers Arch. 316: 34-50.Nattie, E.E. (1983). Ventilation during acute HCl infusion in intact and chemodenervatedconscious rabbits. Respir. Physiol. 54: 97-107.Nattie, E.E. and J.M. Adams (1988). DIDS decreases CSF HCO3 - and increasesbreathing in response to CO2 in awake rabbits. J. Appl. Physiol. 64: 397-403.Nioka, S., B. Chance, M. Hilberman, G.V. Subramanian, J.S. Leigh, Jr., R.L. Veech,and R.E. Forster (1987). Relationship between intracellular pH and energy metabolism indog brain as measured by 31 P-NMR. J. Appl. Physiol. 62: 2094-2102.Nishimura, M., D.C. Johnson, and H. Kazemi (1988). Effects of inhibitors on chlorideoutflux from cerebrospinal fluid. J. Appl. Physiol. 64: 2183-2189.Noll, F. (1974). Determination with LDH, GPT and NAD. In: Methods of EnzymaticAnalysis, Vol. 3, ed. by H.U. Bergmeyer. Academic Press, New York. pp. 1475-1477.Obika, L.F.O. (1989). Urinary kallikrein excretion during potassium chloride infusion inpotassium-adapted rats: effect of amiloride. Clin. Sci. 77: 21-27.8 6Parker, I. and P.H. Zhu (1987). Effects of hypertonic solutions on calcium transients infrog twitch muscle fibres. J. Physiol. (Lond.) 383: 615-627.Patterson, R.W. and S.F. Sullivan (1978). Dererminants of oxygen uptake during sodiumbicarbonate infusion. J. Appl. Physiol. 45: 399-402.Powell, R.L., M.R. Fedde, R.K. Gratz and P. Scheid (1978). Ventilatory response toCO2 in birds. I. Measurements in the unanesthetized duck. Respir. Physiol. 35: 349-359.Rigual, R., J.R. Lopez-Lopez and C. Gonzalez (1991). Release of dopamine andchemorecepor discharge induced by low pH and high PCO2 stimulation of the cat carotidbody. J. Physiol. 433: 519-531.Rigual, R., E. Gonzalez, S. Fidone and C. Gonzalez (1984). Effects of low pH onsynthesis and release of catecholamines in the cat carotid body in vitro. Brain Res. 309:178-181.Roberts, J.K.M., N. Wade-Jardetzky and O. Jardetzky (1981). Intracellular pHmeasurements by 31 13 nuclear magnetic resonance. Influence of factors other than pH on31 1) chemical shifts. Biochem. 20: 5394-5402.Robin, E.D., R.D. Whaley, C.H. Crump, A.G. Bickelmann and D.M. Tracis (1958).Acid-base relations between spinal fluid and arterial blood with special reference tocontrol of ventilation. J. Appl. Physiol. 13: 385-392.Rocher, A., A. Obeso, C. Gonzalez and B. Herreros (1991). Ionic mechanisms for thetransduction of acidic stimuli in rabbit carotid body glomus cells. J. Physiol. 433: 533-548.Roos, A. and W.F. Boron (1981). Intracellular pH. Physiol. Rev. 61: 296-434.Rosenbaum, B.J., D.L. Makoff and M.H. Maxwell (1969). Acid-base and electrolytechanges induced by acute isotonic saline infusion in the nephrectomized dog. J. Lab. Clin.Med. 74: 427-435.Ruch, F.E. and M.R. Hughes (1975). The effects of hypertonic sodium chloride injectionon body water distribution in ducks (Anas platyrhynchos), gulls (Larus glaucescens) androosters (Gallus domesticus). Comp. Biochem. Physiol. 52: 21-28.Saito, K., Honda, Y. and N. Hasumura (1960). Evaluation of respiratory response tochanges in pCO2 and hydrogen ion concentration of arterial blood in rabbits and dogs.Jap. J. Physiol. 10: 634-645.8 7Scheid, P. and J. Piiper (1986). Control of breathing in birds. In: Handbook ofPhysiology, Sect. 3, The Respiratory System, Vol. II, Part 2, ed. by N.S. Cherniack andJ.G. Widdicombe. American Physiological Society, MD. pp. 815-832.Schuitmaker, J.J., A. Berkengosch, J. De Goede and C.N. Olievier (1986). Effects ofCO2 and H+ on the ventilatory response to peripheral chemoreceptor stimulation. Respir.Physiol. 64: 69-79.Shimizu, M. and D.R. Jones (1987). Acid-base balance in ducks (Anas platyrhynchos)during involuntary submergence. Am. J. Physiol. 252: R348-R352.Shires, G.T. and J. Holman (1948). Dilution acidosis. Ann. Int. Med. 28: 557-559.Somero, G.N. (1986). Protons, osmolytes, and fitness of internal milieu for proteinfunction. Am. J. Physiol. 251: R197-R213.Sotos, J.F., P.R. Dodge and N.B. Talbot (1962). Studies in experimental hypertonicity.II. Hypertonicity of body fluids as a cause of acidosis. Pediatrics 30: 180-193.Stephenson, R. and D.R. Jones (1993). Energy metabolism of the Pekin duck duringprolonged head submergence using 31 P-MRS. Am. J. Physiol., in press.Stewart, P.A. (1983). Modern quantitative acid-base chemistry. Can. J. Physiol.Pharmacol. 61: 1444-1461.Urena, J., J. Lopez-Lopez, C. Gonzalez and J. Lopez-Barneo (1989). Ionic currents indispersed chemoreceptor cells of the mammalian carotid body. J. Gen. Physiol. 93: 979-999.Walker, R.L., A.E. Buie and N.L. Armstrong (1990). The effects of saline exposure onthe electrolyte, acid-base and blood oxygen status of the toad, Bufo marinus. Bulletin Can.Soc. Zool. 21: 95.Wathen, R.L., R.A. Ward, G.B. Harding and L.C. Meyer (1982). Acid-base andmetabolic responses to anion infusion in the anesthetized dog. Kidney Int. 21: 592-599.Wesson, D.E. (1990). Depressed distal tubule acidification corrects chloride-depletealkalosis in rats. Am. J. Physiol. 259: F636-F644.Whalley, D.W., P.D. Hemsworth and H.H. Rasmussen (1991). Sodium-hydrogenexchange in guinea-pig ventricular muscle during exposure to hyperosmolar solutions. J.Physiol. (Lond.) 444: 193-212.8 8Wilding, T.J., B.Cheng and A. Roos (1992). pH regulation in adult rat carotid bodyglomus cells. J. Gen. Physiol. 100: 593-608.Wilkes, P.R.H. and B.R. McMahon (1986). Responses of a stenohaline freshwater teleost(Catostomus commersoni) to hypersaline exposure. I. The dependence of plasma pH andbicarbonate concentration on electrolyte regulation. J. exp. Biol. 121: 77-94.Winters, R.W., P.R. Scaglione, G.G. Nahas and M. Verosky (1964). The mechanism ofacidosis produced by hyperosmotic infusions. J. Clin. Invest. 43: 647-658.Yamada, K. (1970). The increase in the rate of heat production of frog's skeletal musclecaused by hypertonic solutions. J. Physiol. (Lond.) 208: 49-64.Zar, J.H. (1984). Biostatistical Analysis. Prentice-Hall, NJ.Zeidler, R. and H.D. Kim (1977). Preferential hemolysis of postnatal calf red cellsinduced by internal alkalinization. J. Gen. Physiol. 70: 385-401.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items