UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Anoxia tolerant hepatocytes : a model system for the study of metabolic and channel arrest Buck, Leslie T. 1993

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

Item Metadata


831-ubc_1993_fall_phd_buck_leslie.pdf [ 4.49MB ]
JSON: 831-1.0086455.json
JSON-LD: 831-1.0086455-ld.json
RDF/XML (Pretty): 831-1.0086455-rdf.xml
RDF/JSON: 831-1.0086455-rdf.json
Turtle: 831-1.0086455-turtle.txt
N-Triples: 831-1.0086455-rdf-ntriples.txt
Original Record: 831-1.0086455-source.json
Full Text

Full Text

Anoxia Tolerant Hepatocytes: A Model System for the Study of Metabolic andChannel Arrest.ByLeslie Thomas BuckB.Sc. Hons. Biology/Chemistry. Brock University, 1986A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Zoology)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAJune 1993©Leslie Thomas Buck 1993In 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 Z-oo Lo 6 yThe University of British ColumbiaVancouver, CanadaDate  frA4-e-^4or3DE-6 (2/88)iiAbstract.Although there are a large number of studies on the adaptive mechanismsby which organisms tolerate and survive hypoxic or anoxic periods, this literatureconsists of experimental information gathered from a heterogeneous assemblageof species at the whole-animal and tissue level of organization, making acomprehensive understanding of the intracellular anoxic defense mechanismsdifficult to obtain. A cell level system that responses to anoxia in a fashion similarto that of the anoxia-tolerant animal from which it was obtained would beinvaluable.Described here is a method for the isolation of anoxia tolerant hepatocytesfrom the freshwater turtle Chrysemys picta bellii. Freshly isolated hepatocyteswere determined to be viable based on trypan blue exclusion, LDH leakage,gluconeogenic capacity from 14C-lactate and responsiveness to adrenalin,glucagon, and insulin, and maintenance of cellular [adenylate]. During 10 h ofanoxic incubation cell staining, LDH leakage, and cellular [ATP] was unchanging,but the rate of ATP turnover decreased by 90%.The microcalorimetrically measured heat flux from hepatocytes insuspension (25°C) decreased by 76% in response to anoxia. To account for thedifference between the two measurements of metabolic rate, the heat flux fromknown anaerobic end products was calculated (using caloric equivalents). Twodominant pathways, known to be functional (glycogen fermentation to lactate andthe breakdown of glycogen to free glucose), accounted for only 42% of the anoxicheat flux, resulting in an "exothermic gap" (58%). This differs from thenormoxically incubated hepatocytes where the indirect calorimetric measurementof heat flux (hepatocyte 02 consumption) could fully account for thecalorimetrically measured heat flux. When normoxic hepatocytes were inhibitediiiwith cyanide a rapid mono-phasic suppression in heat flux was observed,suggesting that there is no short-term Pasteur Effect in these cells.Since [ATP] was unchanging in the presence of a 10 fold decrease in ATPturnover, a concomitant decrease in ATP utilization seemed essential. In view ofthe large fraction of total cellular energy metabolism used to support ion gradientsthe activity of the plasma membrane Na + /K + ATPase and membrane potentialwas measured in response to anoxia. From normoxic hepatocyte suspensions theouabain inhibitable 86Rb + uptake was determined and found to comprise 28% ofthe total normoxic cellular ATP turnover. In response to anoxic incubation theactivity of the pump decreased by 75%; however, this comprised 74% of the totalanoxic ATP turnover. Plasma membrane potential was measured during anoxia,using the distribution of 36CI - , and was not significantly different from thenormoxic measurement. Since the plasma membrane potential was maintainedduring anoxia, and since the activity of the Na + /K + ATPase decreased, the flux ofions across the plasma membrane must also have decreased (a low permeabilitystate termed "channel arrest").Taken together, an anoxia tolerant hepatocyte preparation has beendeveloped that undergoes a 10 fold reduction in ATP turnover in response toanoxia (metabolic arrest). Furthermore, a 4 fold decrease in the rate of ATPutilization by Na + /K + ATPase, and the lack of a change in membrane potential, inresponse to anoxia suggests that mechanisms exist that achieve a functionaldecrease in ATP utilization and membrane permeability (channel arrest). Within acellular system such as this the more complex regulatory mechanisms involved in alarge coordinated reduction in ATP turnover rates can be probed during normoxic -anoxic transitions.ivTable of Contents.Abstract^  iiTable of Contents^ ivList of Tables  viiList of Figures^  viiiAcknowledgements ixChapter 1: The Western Painted Turtle (Chrysemys picta bellii) as a model for theStudy of Anoxia Tolerance Mechanisms.Introduction^  1Good Animal Anaerobes ^  2(1) Storage of Fermentable Substrate^  3(2) Alternative Fermentation Pathways  3(3) Endproduct Accumulation^  6Mechanisms of Metabolic Arrest  9(1) Covalent Modification ^  11(2) Allosteric Regulation  14(3) Reversible Association of Enzymes into Multi-EnzymeComplexes and to Subcellular Components^ 14Mechanisms of Channel Arrest ^  16Anoxia Tolerant vs Anoxia Intolerant Liver Cell Death^ 25Summary^ 28Chapter 2: Preparation and Characterization of Anoxia Tolerant Hepatocytes.Preface ^  30Introduction  30Materials and Methods^  31Hepatocyte Isolation  31Cell Viability^  32VOxygen Consumption Measurement^ 33Enzyme Compartmentation^ 33Gluconeogenic Rate Determination ^ 34Assessment of Anoxia Tolerance 34Conversion Factors ^ 36Chemicals^ 36Statistics 37Results^ 37Assessment of Cell Quality^ 37Recovery Oxygen Consumption 42Determination of Anoxia Tolerance^ 42Anoxic Carbohydrate Metabolism 42Conversion Factors ^ 43Discussion^ 43Chapter 3: Microcalorimetric Measurement of Reversible Metabolic SuppressionInduced by Anoxia.Preface ^  56Introduction  56Materials and Methods ^  57Experimental Animals  57Hepatocyte Preparation^  57Microcalorimetry^  58Oxygen Consumption measurement^  59Chemicals^  59Statistics  59Results^  60Microcalorimetry^  60viCR Ratio Determination^  63Anoxic Heat Flux  63Discussion^  67Chapter 4: Suppression of Na + /K + ATPase Activity and a Constant PlasmaMembrane Potential in Hepatocytes During Anoxia: Evidence in Support of theChannel Arrest Hypothesis.Preface ^  75Introduction  75Materials and Methods^  77Animals and Hepatocyte Preparation^  77Measurement of Na + /K + ATPase Activity  78Membrane Potential Measurement^  80Chemicals^  81Results^  81Ouabain Inhibitable Na + /K + ATPase Activity^ 82Membrane Potential Measurement^  85Discussion^  88Chapter 5: General Discussion^  93Research Summary 93Regulation of a Coordinated Metabolic Suppression ^ 95Pathways Regulating a Coordinated Metabolic Suppression^ 100Literature Cited ^  107viiList of Tables1. Assessment of cell quality following isolation and 24 hours ofincubation at 4°C (storage conditions).^  382. The effect of adrenalin, glucagon, and insulin on the rate ofgluconeogenesis.^  393. Total enzyme activities and the compartmentation of PEPCKand the marker enzymes CS and LDH in the liver ofChrysemys pitta.^  404. Trypan blue exclusion, LDH leakage and energy charge after 10 hoursof anoxic incubation^  415. A comparison between direct and indirect measures of metabolicsuppression. ^  626. Oxygen consumption of hepatocytes at 10 and 25°C, and the ratio ofcalorimetric heat flux to respirometric oxygen flux (CR ratio)^ 647. Proportion of total cellular ATP turnover utilized by the Na + /K + ATPaseunder normoxic and anoxic conditions. ^  848. A comparison of hepatocyte plasma membrane potential under anoxicand normoxic conditions and as effected by highK + and valinomycin medium. ^  87viiiList of Figures.1. A summary of the probable sequence of events leading to cell deathin an anoxia intolerant mammalian liver. ^  272. A comparison between normoxic and anoxic adenylate concentrationsand energy charge in 10 h hepatocyte incubation.^ 453. Glucose production from turtle hepatocytes incubated for 10 h ofnormoxia or anoxia^ 464. Lactate production from turtle hepatocytes incubated for 10 h ofnormoxia or anoxia^  475. Glycogen production from turtle hepatocytes incubated fro 10 h ofnormoxia or anoxia^  486. A simultaneous calorimetric trace of normoxic and anoxic hepatocytes ^ 617. Normoxic and anoxic hepatocyte heat flux measured at 25 and 10°C ^ 668. A calorimetric trace showing the effect of iodoacetate on anoxicheat flux in hepatocytes ^  689. A simultaneous calorimetric trace of normoxic and anoxichepatocytes as effected by dinitrophenol and cyanide^ 6910. Net uptake of 86Rb + by anoxic and normoxic hepatocytes ^ 8311. Plasma membrane potential as determined by the distribution of 36CI - ^ 8612. A schematic diagram of possible control pathways involved ina coordinated down-regulation of energy metabolism. ^ 102ixAcknowledgements.I would like to thank my supervisor Peter Hochachka for his support andenthusiasm throughout my Ph.D. studies. I would also like to thank Dr.T.PMommsen for advice concerning hepatocyte isolation techniques and theradioisotopic measurement of gluconeogenic rates; Dr. W.K. Milson for the use ofa Radiometer 02 electrode and meter; and Drs. R.K. Suarez, C.D. Moyes, T.G.West C.J. Doll, P.M. Schulte, S.C. Land, and J.F. Staples for many helpfuldiscussions. Also, Thomas Haller, Martin Ortner, and Karel Cerny for helpfuldiscussions and assistance with equipment during the calorimetry study, and Drs.T. Glass, J. Mehroke and B. Malhotra for use of laboratory space and helpfuldiscussions concerning the 86Rb + experiments.Special thanks to my wife Liana for her support, tolerance, and patienceduring my studies; and my son Thomas for lots of playtime fun.1Chapter 1. The western painted turtle (Chrysemys picta bellii) as a model for thestudy of anoxia tolerance mechanisms.IntroductionOver the last decade the strategies by which invertebrates and ectothermicvertebrates tolerate and survive oxygen limiting conditions have become betterunderstood. It is evident from studies of anoxia tolerant species that it is anintegrated metabolic response, whereby both energy producing and energyutilizing pathways are down-regulated in a coordinate fashion that permits long-term anoxic survival. Importantly, it appears to be the lack of such a responsethat commits endotherms to being anoxia sensitive. Anoxia is rapidly fatal to mostmammals, and it is only the diving members of this group that can tolerate oxygenlimiting conditions for 1-2 hours (Hindell et al. 1992), which is still aninconsequential period of time in comparison to good animal anaerobes. Only alimited amount of information regarding naturally-evolved defense mechanisms bywhich anoxia tolerance can be improved can be gained using anoxia intolerantorganisms. Although a large literature has developed around protecting hypoxiasensitive systems against hypoxia or anoxia, (Brezis et al. 1984; Herman et al.1990; Gregory et al. 1990; Lemasters et al. 1993; Hansen 1985) the standard(usually mammalian systems) in such studies suffer from a serious limitation: thelack of a naturally-evolved hypoxic defense mechanism. For this purpose it seemsessential to turn the study of anoxic/hypoxic defense mechanisms employed bygood animal anaerobes: helminths, intertidal bivalves, crucian carp and goldfish,and perhaps the most anoxic tolerant vertebrate of all - the freshwater turtle.The literature regarding the metabolic adaptations of diving and hibernatingmammals, ectothermic vertebrates such as goldfish, carp and freshwater turtles,2and invertebrates to hypoxic or anoxic conditions has been extensively reviewedelsewhere: Storey and Storey 1990; Lutz 1992; Hochachka and Somero 1984;Hochachka 1986; Hochochka and Guppy 1987; van Waarde 1990; van denThillart 1985; de Zwaan 1983. Therefore, in this review of the literature particularattention will be focused on the adaptive mechanisms of freshwater turtles andespecially Chrysemys picta bellii, since it is the species from which the anoxiatolerant hepatocyte model presented in this dissertation was developed. However,much of the data regarding adaptations to anoxia comes from the study ofinvertebrates or other ectothermic vertebrates, and these data will be referred towhere the specific adaptation is better represented. The fact that these data comefrom a heterogenous and diverse group of organisms is a limitation in itself, sinceany proposed anoxic defense mechanisms are necessarily composites from studiesof these diverse groups.In the following literature review I will discuss the metabolic adaptations ofgood animal anaerobes and in particular those of the western painted turtleChrysemys picta bellii. Furthermore, what is likely the most effective long-termstrategy of all, the coordinated reduction of both energy production and utilization,processes encompassed by two unifying hypoxia defense hypotheses - metabolicarrest and channel arrest - will be discussed.Good Animal AnaerobesGood animal anaerobes have been characterized by several metabolicadaptations that increase anoxic survival, and these include: 1) an increasedendogenous storage capacity of fermentable substrates, 2) the use of alternativefermentation pathways to increase the anaerobic ATP yield to reduce substrateusage and endproduct accumulation, 3) enhanced buffering capacity and3compensatory ion changes to minimize the effects of high concentrations ofanaerobic endproducts, and 4) metabolic rate depression (Hochackha 1985, 1986;Storey and Storey 1990). The last of these, a down-regulation of metabolism,potentially offers the most profound extension of anoxic survival time.1) Storage of fermentable substrate. In order to survive oxygen limitingconditions it is essential to have sufficiently high stores of fermentable substrateto last the duration of the anoxic episode. Glycogen storage in the liver is by farthe most important storage site during anoxia. Glycogen concentrations in theliver of hypoxia tolerant species such as the goldfish and turtle reach levels as highas 1300 and 860 ymol glucosyl unit.g -1 , respectively. This is 4 to 5 fold in excessof the concentrations measured in intolerant species such the rat and trout of 210and 235 pmol glucosyl unit.g -1 , respectively. Moreover, non-central glycogenstores including heart and skeletal muscle are also in excess of those measured inhypoxia sensitive animals (Hochachka and Somero 1984).2) Alternative fermentation pathways. In addition to increased stores offermentable substrate, good animal anaerobes employ an array of alternativefermentative pathways and endproducts. These pathways are more efficient, interms of ATP produced per glucose utilized, than the more commonly thought ofpathway of glucose to lactate fermentation. Glucose to propionate fermentation,for example, can yield up to 6 mol ATP. mol glucose -1 , whereas fermentation tolactate yields 1 ATP per lactate produced. Another advantage of alternativeendproducts and pathways is the lower degree of acidification offered. Forglucose (glycogen) to lactate fermentation, assuming 1:1 coupling with cellularATPases, 1 (1.5) ATP are cycled through per ,umol H + accumulated. For theglucose to succinate or propionate fermentations 2 or 3 pmol ATP are cycled per4mol H + accumulated, respectively (Hochachka and Somero 1984). Additionally,fermentation to succinate and/or proprionate involve mitochondrial pathways,producing extra ATP and also generating a partial mitochondrial electrochemicalgradient, which could be an important factor in sustained anoxic survival(Holwerda and de Zwaan 1979). Glucose fermentation to octopine, lysopine,alanopine, strombine, succinate or proprionate are known to occur in variousbivalve molluscs (de Zwaan 1983; Hochachka et al. 1983). The most importantfermentation pathway in C. picta, apparent from the accumulation to highconcentrations during anoxia, is glycolysis to lactate (Jackson et al. 1982; Buckand Hochachka 1993). That C. picta relies on glycolytically derived fermentationwas demonstrated by its survival under an atmosphere of N2 and its subsequentdeath when a subcutaneous injection of iodoacetate (a glycolytic inhibitor) wasadministered (Belkin 1962). For a more thorough treatment of alternativefermentation pathways in a variety of species see Hochachka and Somero (1984).Calorimetry provides a direct method of determining if an organism utilizesalternative endproduct pathways. Through the use of established caloricequivalents (heat produced per mole substrate catabolized in kJ.mol substrate -1 ,Gnaiger 1983) for a given substrate it is possible to account for the measured heatproduction. The difference then, between the total heat dissipation measuredduring anoxia and that calculated from biochemical endproduct analysis, would besuggestive of additional anoxic reaction and alternative endproducts. For oxidativemetabolism in general, all of the heat produced can be accounted for from knownbiochemical processes (Gnaiger, 1983). Since this is the case, an indirectestimate of heat production can be calculated from a measure of oxygenconsumption and an oxycaloric equivalent (heat produced per mole oxygenutilized, kJ.mol 02 -1 , Gnaiger, 1983).5Although under aerobic conditions there is a good correlation betweenmetabolic heat produced and oxygen consumed, this is not the case underanaerobic conditions. The measured metabolic heat produced cannot beaccounted for by the known fermentation pathways. This was first identified in anaccounting of anoxic metabolic heat sources in aquatic oligochaetes and mussels(Gnaiger 1980a, 1980b). After accounting for the heat produced by metabolicendproduct accumulation, only 50% of the heat measured in the calorimeter couldbe accounted for. This work has been criticized by Shick et al. (1983), pointingout that the endproduct analysis was not performed on the same animals used inthe calorimetric study (numbers were taken from the literature), furthermoreseasonality was not considered. When this study was repeated by Shick et al.(1983) on Mytilus edulis, carefully measuring metabolic heat production andmetabolic endproducts of the same animals, they confirmed the earlier finding byGnaiger (1980a, 1980b). They measured an "anoxic gap" (unaccounted formetabolic heat) on the order of 40 - 60%, similar to the value estimated byGnaiger. More recently Hardewig et al. (1991) have been able to account for upto 83% of the heat production through endproduct analysis and careful accountingof biological buffers in the marine worm Sipunculus nudus.Thus far the source of the missing heat component is unknown. This islikely due to the use of whole-animal systems, making the identification of specificheat sources more difficult. In a cellular system where extracellular variables canbe controlled, such as the anoxia tolerant hepatocyte preparation described inChapter 2, metabolic heat sources should be more easily indentified.3) Endproduct accumulation. With reliance on anoxic fermentation comes theassociated toxicity from accumulation of large concentrations of anionic, organicendproducts and decreasing pH. The problem of large concentrations of6endproducts and charge imbalance is partially overcome by movement of counter-ions, such as Ca2+ and Mg2+ , into the blood space. Counter-ion movement intothe blood space has been shown to occur in bivalve molluscs and turtles. In theseanimals there is a good correlation between blood counter-ion and endproductaccumulation, the counter-ions probably released form bone and/or shell stores (deZwaan 1983; Jackson 1993).The most dramatic increase in plasma lactate concentrations so far observedin vertebrates is in the fresh water turtle (C. picta), where levels as high as 200mM have been measured after long-term anoxic submergence at 3°C (Ultsch andJackson 1981). Over a 12 week anoxic dive, where plasma lactateconcencentration increased to over 100 mM, plasma counter-ion concentrations(Na t , K .', Mg 2+ , Ca2+ increase and HCO3 - , Cl - decrease) were found to balancethat of lactate (Jackson and Heisler 1982). As well as balancing charge, the intraand extracellular spaces are also osmotically balanced, indicated by the distributionof water between intra and extracellular compartments remaining constant duringanoxia (Jackson and Heisler 1983). Osmotic balance is achieved by the similaraccumulation of solutes intracellularly as extracellularly. However, the measuredconcentrations of intracellular solutes could not fully account for the increase inosmolality, and it is likely that other unmeasured solutes accumulate (Jackson andHeisler 1983). Turtles do not have an exceptionally high blood buffering capacity(Ultsch et al. 1984), nor is tissue buffering capacity exceptionally high, 50, 97,and 118 slykes for skeletal muscle, liver, and heart, respectively (Wasser et al.1991) in comparison to mammals or invertebrates where it has been measured(Castellini and Somero 1981; Morris and Baldwin 1984). However, an importantbuffering mechanism has been shown to be bicarbonate titration, HCO3 - beingreleased form shell stores during anoxia (Jackson 1993). Blood Pc02 levels have7been found to increase to as high as 30-60 mmHg during long-term anoxia in C.picta (Ultsch et al. 1984).From the measured concentrations of fermentable substrate stores(glycogen) in the livers of various animals, mentioned above, Hochachka (1986)calculates that if anaerobic ATP turnover were to be maintained at the same rateas aerobic turnover, the anoxic episode could only be sustained for 3-4 days.Therefore, the above adaptations (1) - (3) are of limited value in extending anoxicsurvival time and could be considered short-term adaptations, whereas (4) theability to depress metabolism, alone or in concert with the first three options, hasthe potential to profoundly extend anoxic survival time and could be considered along-term adaptation strategy. When the extent of metabolic depression isfactored in anoxic survival time is greatly increased. For example in the goldfishsurvival time is increased 5-fold, for the intertidal bivalve Mytilus sp. 20-fold, andfor the turtle, Chrysemys picta bellii, an increase of over 60-fold (Hochachka1986). Obviously, a suppression of metabolism will reduce the impact of highlevels of fermentative endproducts and acidification and possibly, in extremesituations, may reduce the need for other adaptations, such as more efficientfermentative pathways.For a vertebrate the metabolic suppression observed in C. picta in responseto anoxia is exceptional. Jackson (1968) has shown, using whole animalcalorimetry, that the turtle can reversibly undergo an 85% decrease in heatdissipation during anoxic submergence (at 24°C). More recently a greater than99% reduction in metabolism, in response to the combined effects of anoxia anddecreasing, temperature has been demonstrated (20-3°C, Herbert and Jackson1985).The metabolic depression observed in good animal anaerobes such as theturtle has been termed "metabolic arrest" (Hochachka 1986) and refers specifically8to an active process whereby metabolism is switched down, or in the extremecase of the brine shrimp, apparently off (Busa 1985; Storey and Storey 1990).This fundamentally differs from the response to anoxia of hypoxia sensitiveorganisms such as mammals. When faced with oxygen limiting conditionsmammals typically undergo a glycolytic activation or Pasteur Effect to compensatefor the loss of oxidative ATP synthesis (Pasteur Effect is defined as an increase inthe rate of carbon flux through glycolysis in response to oxygen limitingconditions). There is an apparent lack of this response in more efficient animalanaerobes. There are several mechanisms by which this could be achieved, andthese will be discussed below. By implication it has been suggested that the rateof energy utilization, in terms of ATP units, must also be down-regulated(Hochachka 1986). However, the evidence thus far has been indirect, and a directmeasure of a functional decrease in the activity of an ATP consuming reactionshas not been demonstrated. It is clear; however, that in response to a singleenvironmental signal (such as oxygen limitation) a mechanism that simultaneouslyreduces both the rate of ATP synthesis and utilization is required. The responsemust also be rapid and reversible when oxygen is no longer limiting.It has been suggested that, upon the onset of anoxia, anoxia tolerantorganisms undergo a short-term activation of glycolysis, compensating for theimmediate ATP shortfall due to the loss of oxidative metabolism. Subsequently,the glycolytic activation would be reversed (a reversed Pasteur Effect, Hochachka1986), reducing glycolysis to a lower level that would be sustained on a long-termbasis. The issue as to whether C. picta undergoes a short-term glycolyticactivation in response to anoxia is not settled. On an organ by organ basis, itseems that turtle brain does experience a short-term glycolytic activation (Robin etal. 1979, Lutz et al. 1984 and 1985; Kelly and Storey 1988; Clark and Miller1973). The data are not as convincing for turtle heart. In one case a decrease in9tissue lactate was recorded (Kelly and Storey 1988), perhaps the result of a wellperfused tissue, and in another case tissue lactate concentration had to beestimated from an overlapping peak on a 1 H nuclear magnetic reasonance (NMR)spectrum. Changes in the concentration of the confounding compound due toanoxia were not considered, and furthermore, the rate of lactate accumulationover the anoxic period shown in the publication appears linear. This makes adistinction between a rapid rate of lactate accumulation (Pasteur Effect) and asubsequent transition to a lower rate difficult to establish (Wasser et al. 1992).From direct measures of lactate accumulation over a 5 h time course a short-termglycolytic activation within the first hour is observed in the liver (Kelly and Storey1988). However, these measures come from force dived animals, and there is thepossibility that the initial rapid rate of lactate accumulation is a stress response.Mechanisms of Metabolic ArrestAs mentioned above, hypoxia sensitive organisms, or the technically moreeasily evaluated responses of perfused organs and cells from such organisms,typically demonstrate a glycolytic activation (Pasteur Effect) in face of oxygendeprivation. When fully oxidized, glucose yields 36 ATP (36 mol ATP/ molglucose), whereas glucose fermentation to lactate yields only 2 mol ATP/ molglucose, an energetic shortfall on the order of 18 fold. A glycolytic activation ofthis magnitude has never been observed (in mammalian tissue at least). Forexample, rat liver cells display a 7-fold activation (Hue 1982), while the largestactivation of glucose flux through glycolysis was 14 fold, observed in bull sperm(Hammerstedt and Lardy 1983). Moreover, mammalian brain, one of the mostanoxia sensitive tissues, can only activate glycolysis by 9-10 fold over a shortperiod (6-10 min), rapidly declining thereafter (Kinter et al. 1984). In anoxia10sensitive organisms and tissues as above, the loss of ATP generating capacity byoxidative phosphorylation is never fully compensated for by activated glycolysis.The strategy of increasing glycolytic flux to make up for the ATP shortfallcaused by the loss of oxidative metabolism in anoxia is clearly not the optimalapproach for long-term sustained anoxic survival. As discussed earlier, goodanimal anaerobes lower metabolic rate and thus reduce energy requirements duringconditions of oxygen deprivation. Data from calorimetry and biochemicalendproduct analysis studies are available for various species and tissue-levelexperiments. These include, freshwater turtles (C.picta, Pseudemys scripta)studies at the whole-animal level: Jackson 1968, Jackson and Ultsch 1982,Herbert and Jackson 1985a & b, Ultsch et al. 1984. Studies at the tissue level:brain, Lutz et al. 1985, Lutz 1992, Robin et al 1979. In other species: theintertidal bivalve (M. edulis) Schick 1983, and the goldfish van Waversveld et al.1989. All the above undergo reversible metabolic suppression to levels 5-10, 5-20, and 20-33% of resting normoxic rates, respectively. It therefore follows thata coordinated metabolic suppression is occurring in species tolerant of long-termanoxia. Thus the question - how is a reduction of this magnitude achieved?Several mechanisms by which glycolytic flux could be switched down havebeen described (Hochachka 1985; Storey and Storey 1990) and include: (1)covalent modification of key glycolytic enzymes by phosphorylation -dephosphorylation cycles activated by Ca 2+ ; (2) allosteric regulation of keyglycolytic enzymes, such as hexokinase (HK) inhibition by glucose -6- phosphate(G6P) and the decreased concentration of a potent phosphofructokinase (PFK)activator - fructose -2,6- bisphosphate (F-2,6-P2); and (3) enzyme or pathwaycompartmentalization, either by the reversible association of functional multi-enzyme complexes or binding of glycolytic pathway enzymes to cytoskeletalcomponents or myofibrils.11The most comprehensive analysis of anoxia-induced metabolic arrestmechanisms has been performed for the channeled whelk (for review see Storeyand Storey 1990). Similar information is emerging from studies on the turtle. Thedata from both species will be discussed below.1) Covalent modification. The reversible phosphorylation of an enzyme inresponse to anoxia was first demonstrated for pyruvate kinase (PK) from theadductor muscle of the intertidal bivalve M. edulis (Siebenaller 1979; Holwerda etal. 1983). Pyruvate kinase catalyzes the conversion of phosphoenol-pyruvate(PEP) to pyruvate and is considered one of the rate limiting enzymes of glycolysis.Therefore, an inhibition at this point in the pathway could be an important factor indown-regulating flux through the entire pathway. Kinetic analysis of the effects ofwhelk red muscle PK phosphorylation has shown that the maximum activity of theenzyme decreased 3 fold, the half-saturating substrate concentration (PEP)increased 12-fold, the half-saturating activator concentration (fructose 1,6-bisphosphate (F1,6P2)) increased 24-fold, and the half-saturating inhibitorconcentration (Mg-ATP) decreased 3.5-fold while that of L-alanine (a potentinhibitor of PK) decreased 490-fold (Plaxton and Storey 1984). The net result ofphosphorylation then, is an enzyme that has reduced maximal activity, hasreduced affinity for its substrate, is less responsive to allosteric activation and isstrongly inhibited by Mg-ATP and L-alanine. The strong inhibition by L-alanine isparticularly important since it accumulates as an anaerobic endproduct in thisspecies. The evidence for phosphorylation of this enzyme is very good, since asStorey and Storey (1990) point out: (1) 32P is readily incorporated into the enzymein vivo under anoxia, but not under aerobic conditions, (2) alkaline phosphatasetreatment of the anoxic enzyme form to remove the covalently bound phosphaterestores the aerobic kinetic properties of the enzyme, and (3) in vitro incubation of12cell-free homogenates from aerobic tissues with Mg 2+ and ATP converts aerobicPK to the anoxic form (Plaxton and Storey 1984). Additionally, three otherenzymes important in the regulation of glycolysis, were also found to undergoreversible phosphorylation. These include glycogen phosphorylase (GP), 6-phosphofructo-1-kinase (PFK-1) and 6-phosphofructo-2-kinase (PFK-2) fromvarious whelk organs. In each case the maximum activity of the enzyme waslowered, and kinetic constants change to produce a less active enzyme (Storeyand Storey 1990).A similar pattern of phosphorylation control of key glycolytic enzymes inresponse to anoxia is arising from studies of two vertebrates, goldfish (Storey1987) and fresh water turtles (Brooks and Storey 1989). Similar results wereobtained from both species; therefore, only the turtle will be considered here. Theactivity of GP, PFK-1 and PK from liver, heart, red, and white muscle and thetissue concentration of F-2,6-P was assessed in aerobic and 5 h-anoxically divedturtles. The important changes as a result of anoxia-induced phosphorylation areas follows: 1) for liver - 3.3 fold increase in the active form of GP, 1.5 foldincrease in the 150 of PFK for citrate, 1.5 fold increase in the K m of PFK for F-6-P,and a 1.2 fold increase in the Vmax of PK; 2) for heart - 2.6 fold decrease in theactive form of GP, 1.7 fold decrease in the K m of PFK for ATP, and 2 fold increasein the 150 of PFK for citrate; and 3) for white muscle - 5 fold decrease in the K m ofPFK for F-6-P, 2 fold increase in the K m of PK for ADP, and 1.7 fold decrease inthe 150 of PK for alanine. For red muscle the only significant change was a 1.5fold increase in the Km value of PFK for ATP. The phosphorylation status of theseenzymes in response to anoxia was not tested. However, the anoxia-inducedkinetic changes are consistent with those observed from kinetic analysis of theanoxic whelk enzymes (above). The tissue specific responses are not universal,but may reflect the role individual tissues play during anoxia. For example, in liver13the active form of GP increases in response to anoxia while the kineticmodifications of PFK suggest a decrease in activity. This is consistent with therole of the liver as a glycogen storage site supplying substrate to other workingtissues during anoxia. In heart the active form of GP decreases, suggesting aconservation of fermentable substrate. Again the kinetic alterations to PFK aresuggestive of a decrease in glycolysis, perhaps in preparation for sustained workby the heart during anoxia. The tissue specific responses of red and white muscleare more varied and difficult to interpret. Red and white muscle show no changesin the active form of GP. Red muscle demonstrated a 1.5 fold increase in the K mof PFK for ATP and a 30% decrease in the V max of PK. While white muscleshowed a 5 fold decrease in the Km of PFK for F-6-P, a 2 fold increase in the K mof PK for ADP, and a 1.7 fold decrease in the 1 50 of PK for alanine (Brooks andStorey 1989). Although the alterations in the kinetic parameters of GP, PFK, andPK are each individually quantitative, the overall effect on glycolytic carbon fluxduring anoxia can only be qualitatively assessed.Recently a cyclic guanosine monophosphate (cGMP) dependent proteinkinase, that stimulates the phosphorylation of pyruvate kinase, has been identified(Brooks and Storey 1990). It was identified in crude homogenates of heart,radular retractor, and foot muscle from the whelk Busycon caniculatum byexamining changes in the kinetic parameters (as above) of PK in response tophosphorylation. The protein kinase stimulators - cyclic adenosine monophosphateand Ca t + plus phorbol esters did not stimulate phosphorylation; however, theaddition of cGMP to the homogenate increased the rate of phosphorylation by 3-4fold.Since there is evidence for ion channel regulation via G proteins (discussedbelow) the potential for glycolytic carbon flux regulation by cGMP dependentprotein kinases and likely G proteins is exciting. Mechanistically, this would14provide the "missing link" between the coordinate down-regulation of ATPsynthesis and utilization.2) Allosteric Regulation. An important factor in the regulation of carbohydratemetabolism, glucose catabolism via glycolysis versus glucose synthesis bygluconeogenesis, has been shown to be fructose-2,6-bisphosphate (Hue and Rider1987).^F-2,6-bisphosphate activates PFK-1^and inhibits fructose-1,6-bisphosphatase, thus stimulating glycolysis. Through hormonal stimulation,mediated by intracellular second messengers, PFK-2 is phosphorylated andinhibited, decreasing the concentration of F-2,6-bisphosphate, thereby reducingthe stimulatory effect on PFK-1. In whelk organs PFK-2 activity drops 2 - 4 fold inresponse to anoxia and F-2,6-P concentrations drop 4 - 224 fold depending on thetissue (Bosca and Storey 1991). The result of a large decrease in F-2,6-Pconcentrations during anoxia is that gluconeogenesis is switched off and PFK-1 isallosterically regulated to a greater extent by AMP and ATP.Similar changes in the levels of F-2,6-bisphosphate have also been observedin liver, heart, red and white muscle of the turtle in response to anoxia (Brooks andStorey 1989). Liver and white muscle showed no significant change in F-2,6-Plevels in response to anoxia. Although statistically significant, the anoxia induceddecrease in [F-2,6-P] in the heart is minor (approx. 30%) in comparison to thestatistically insignificant (almost 50%) increase observed in white muscle.Moreover, red muscle exhibited a large 200% increase in [F-2,6-P]. Because ofthe variability in these measurements no possible functional explanation was givenby the authors.3) Reversible association of enzymes into multi-enzyme complexes and tosubcellular components. The regulation of enzyme catalyzed reactions or enyzmes15in series - a metabolic pathway, is known to occur at many levels within thecellular environment. The most obvious and extensively studied are the kineticparameters (Km, Vmax) of individual enzymes as influenced by covalent orallosteric modulators and feedback or feedforward modulation (as discussedabove). It is also recognized that enzymes and enzyme pathways can bereversibility bound to membranes, glycogen particles, muscle myofilaments or tomicrofilaments of the cytoskelelaton see Srere 1987 for review). There is noquestion that associations do occur, but the functional significance to overallcellular metabolism is unclear. This is a result of conflicting kinetic analyses ofbound enzymes which indicate that binding can inhibit or stimulate differentenzymes in the same pathway. In fact, the experimental protocol demonstratingenzyme associations and interpretations has been criticized (Brooks and Storey1993).By far the strongest evidence against functional in vivo enzyme-particulateassociations are the isosmotic, low ionic strength isolation conditions typicallyemployed. Under these conditions enzymes can be forced to associate with thecellular particulate fraction. In vitro, when the ionic strength is increased to thatof in vivo cellular levels, no bound enzyme is found; furthermore bound enzyme atlow ionic strength can be made to dissociate from particulate matter by increasingionic strength to in vivo levels (Brooks and Storey 1993). The change in enzymebinding has been correlated with changes in pathway flux, increased binding -increased flux, but the results are not always clear, thus making it more importantto understand what effects binding has on the kinetic parameters of the particularbound enzyme.The binding of glycolytic enzymes to the particulate fraction has beeninvestigated in turtle tissues in response to anoxia. In ventricle muscle thepercentage binding of eight glycolytic enzymes with the particulate fraction16decreases in response to anoxia, and this correlated with a reduced metabolic rate(Plaxton and Storey 1986). However, when the binding of glycolytic enyzmeswas investigated in turtle brain, an increase in binding to the particulate fractionwas observed (Duncan and Storey 1992). It seems that there is a good correlationbetween glycolytic enzyme binding and the activation of glycolysis, since theseresults are consistent with the purported increase in glucose to lactatefermentation in the anoxic turtle brain (Lutz et al. 1985; Lutz 1992). However, akinetic analysis of several bound glycolytic enzymes indicates that binding inhibitsflux of substrate not stimulates flux, with one exception PFK-1 Only PFKdemonstrates an activation upon binding both to F-actin and to band 3 (see Brooksand Storey 1993). Therefore, the correlation between an increase in glycolyticenzyme binding and increased glycolytic carbon flux is perplexing. One of the fewstudies of functional outcomes of enzyme binding to band 3 has clearly shownthat the transition from bound to unbound stimulates lactate production inerythrocytes (Harrison et al. 1991). It still remains somewhat unclear as to thesignificance of enzyme binding to the cellular particulate fraction, but availableevidence points to a role in the down-regulation of energy metabolism in responseto anoxia. Just as a functional role of enzyme binding has been demonstrated inerythrocytes, because of the simplicity and convenience of the cellular model, theturtle hepatocyte model is an ideal experimental model to investigate functionalroles of enzyme binding in response to anoxia.Mechanisms of Channel ArrestThe above mechanisms of glycolytic down-regulation are of obviousimportance in establishing cellular conditions for long-term reliance on glycogen17fermentation. However, as discussed above, a decrease in ATP synthesis mustnecessarily result in a decrease in its utilization.A requisite function of all cell types is the maintenance of ion gradientsacross the plasma membrane lipid bilayer. The transcellular ion gradient is coupledto the formation of a membrane potential for depolarization and repolarization, forsignal transduction, for cell volume regulation, for transport of glucose and aminoacids against their concentrations gradients, for co-transport and counter-transportof ions across the cell membrane and for other transepithelial transport inintestine, kidney, and secretory glands (Skou 1992). This is achieved primarily bythe translocation of 3 Na + out of the cell and 2 K + into the cell using the Na + /K +ATPase. The energy for this process is derived from ATP hydrolysis (Rossier et al.1987). Therefore, Na + /K + ATPase activity and maintenance of cellular gradientsrequires a large fraction of the total cellular demand for ATP. In mammalianhepatocytes the proportion of cellular ATP turnover used to support Na + /K +ATPase activity has been estimated to be between 10 to 40% of ATP turnover,dependent on the method of measurement (Nobes et al. 1989). In mammalianbrain, a tissue where ion transport is a major cellular function, ion pumping mayconsume as much as 50 -60% of the cellular energy expenditure (Erecinska andSilver 1989). The tissue with the largest fraction of energy metabolism partitionedto ion pumping is the kidney, where as much as 80% of the total cellular ATPturnover can be utilized by Na + /K + ATPase activity (Brezis et al. 1984).The large fraction of total cellular energy metabolism used by ion pumps toestablish and maintain ion gradients has been recognized as a possible limitation toanoxic survival (Hochachka 1986). In order to reduce this energy drain duringanoxia it has been hypothesized that membrane permeability to ions (via ionchannels) should be less in anoxia tolerant species. This was coined the channelarrest hypothesis (Hochachka 1986), and the hypothesis makes two predictions:181) that anoxia tolerant cells are less permeable to ions, and 2) that there arefurther decreases in permeability in response to anoxia. The implication is that areduction in the capacity for ions to flow down concentration gradients wouldresult in a significant decrease in the energy required for gradient maintenance. Byinference then, during oxygen limiting conditions the cellular energy demand forion pumping would be decreased and the need for fermentable substrate andmechanisms to deal with anaerobic endproducts lessened.Since all ion channels described thus far are gated (voltage or ligand) insome fashion and ion channels of a cell at rest do not leak ions at significant rates(Hille 1992), is it necessary to reduce membrane ion permeability further to surviveoxygen limiting conditions? More specifically: 1) how much energy is required tomaintain cellular ion gradients at rest, and 2) is there an inherent differencebetween the resting energy expenditure of anoxia tolerant ectothermic vertebratesand the intolerant endothermic vertebrates - mammals? To survive oxygen limitingconditions perhaps all that is needed is a reduction in work, such as a reduction in(1) neuronal action potentials, (2) heart rate, (3) urine concentration by the kidney,and (4) membrane transport of the liver related to its role as a central fuel depotand detoxifying organ. This is a misleading argument since membranes do leakions and there are endo/ectothermic resting cell permeability differences. Ionleakage arises because even at rest there is a low statistical probability of anygiven ion channel being open (Hille 1992, reference to single ion channels), butthere are many different ion channels in any given cell membrane. It is thereforethe summation of the statistical "open state" probabilities of all the particular ionchannels being open at any given moment that gives rise to significant membraneion leaks. Else and Hulbert (1987) have shown that for a given mitochondrialmembrane surface area endotherms have 4-5 times more protein than ectotherms.This is suggestive of a "leakier" mitochondrial membrane in endotherms than in19ectotherms, and it is not unreasonable to assume that this holds true for theplasma membrane as well. In fact, the difference in passive permeability of theliver, kidney, and brain plasma membranes to Na + and K + has been estimatedbetween endotherm and ectotherms (Else and Hulbert 1987). Mammalian(endotherm) membranes were found to be 3-8 fold more permeable to these ionsthan the reptilian homologue. Else and Hulbert (1987) have also demonstratedthat the ouabain inhibitable uptake of 42K+ (an estimate of Na + /K + ATPaseactivity) in liver, kidney, and brain of mammalian tissues used 3-6 fold moreenergy than in reptile tissues. These values also correlate with 4-5 fold greatertissue 02 consumption rates in the mammalian tissues than in reptile tissues,resulting in the suggestion that it is this passive permeability to ions thatcontributes to a large fraction of resting ATP turnover.Even though reptiles inherently have less ion permeable cell membranesthan mammalian cell membranes, they also have a 4-5 fold lower resting ATPturnover. Therefore, to undergo an 85 - 96% (temperature dependent) reductionin resting metabolism ion leaks must in some way be down-regulated (channelarrest). How are ion channels regulated to achieve this?Perhaps one of the most cogent demonstrations of the potential ATPsavings brought about by reduced ion pumping comes from studies of rat andturtle synaptosomes (Edwards et al. 1989). Aerobic incubation of synaptosomesin Na + -free medium or in the presence of ouabain (specific Na + /K + ATPaseinhibitor) reduced oxygen consumption by about 50% in the turtle and about 80%in the rat. The greater inhibition of oxygen consumption in rat synaptosomespossibly was related to a greater Na + channel density in rat brain (Edwards et al.1989). Synaptosomal energy consumption could also be modulated, closing Na +channels with tetrodotoxin substantially decreased oxygen consumption;20conversely increasing channel activity with veratridine substantially increasedoxygen consumption.There is conflicting evidence with regard to the presence of a channel arrestmechanism in turtle brain. It has recently been shown that in response to anoxianeuronal Na + channel density decreases by 42% (Perez-Pinzon et al. 1992).Moreover, after anoxic incubation turtle brain and brain slices have been shown tobe less leaky to K + [during inhibition of the Na + /K + ATPase with ouabain (Chih etal. 1989)] and Ca t + [during glycolytic inhibition with iodoacetate (Bickler 1992)]respectively. In association with a decrease in membrane Na + channels and adecrease in membrane K + and Ca2+ leaks a decrease in membrane conductancewould be expected. However, when neuronal membrane conductance wasmeasured in anoxia, no changes were found (Doll et al. 1989). Recently it hasbeen shown that pre-exposure of turtles to anoxia for 14-18 h reduced theglutamate-mediated Ca 2+ influx in anoxic brain cortical slices in comparison tocortical slices made acutely anoxic (Bickler 1993). This indicates that duringprolonged anoxia the post-synaptic glutamate receptors (N-methyl-D-aspartatereceptors) are down-regulated in some fashion (Bickler 1993). Moreover, proteinkinase C (PKC) inhibitors and stimulators changed the Ca 2+ flux in anoxic slices,suggesting that the receptor may be under phosphorylation control. Thediscrepancy between the evidence pointing to a decrease in membrane ionpermeability and the lack of conductance changes across the cell membranecannot be resolved at present. But the discrepancy suggests that there may be atime dependence to the brain anoxic defense mechanism.It has been suggested that a reduction in electrical activity, suppression ofevoked potential, and increased action potential thresholds (Perez-Pinzon et al.1992; Feng et al. 1989) combined with an inherently less leaky membrane inectotherms (above) will sufficiently reduce the energy requirements of the anoxic21turtle brain to permit long-term anoxia survival (Lutz 1992). This has been termed"spike arrest" by Sick et al. (1993). Whether or not glycolysis alone can sustainthe anoxic energy requirements of the turtle brain is unclear, since themeasurement of anoxic lactate accumulation vary by as much as 10 fold. Theseanoxic rates vary from 50.4 (Robin et al. 1979) to 5.7 pmol ATP.g" 1 .h1 1 (Lutz etal. 1984). However, in light of the evidence (above) suggesting there is areduction in membrane ion permeability, this notion cannot be dismissed.Furthermore, Sick et al. (1993) made reference to an article in press demonstratingthat blockage of the cell membrane adenosine receptor (A 1 ) during anoxia inducedthe rapid loss of transcellular ion gradients, again supporting a channel arrest typemechanism in turtle brain. The A1 receptor has also been shown to interact withan intracellular G protein that interacts directly with a membrane cation channel;when the G protein is in contact with the ion channel cation flux decreases(Rudolphi et al. 1992).Just as some soluble enzymes are regulated by phosphorylation anddephosphorylation events, it is becoming clear that this is also true for membraneion channels (Brown 1991). Recall the activation mechanism of glycogenbreakdown: adrenergic activation of a peripheral membrane G s protein, whichstimulates adenylate cyclase (AC), producing cyclic adenosine monophosphate(cAMP) and subsequent stimulation of cAMP-dependent protein kinase (cAMP-dPKor PKA). Ultimately this signalling cascade phosphorylates and activates glycogenbreakdown by glycogen phosphorylase and phosphorylates and inhibits glycogensynthase (Sutherland 1971). A similar type of signalling mechanism has now beenelucidated for the regulation of membrane ion channel proteins (for review seeBrown 1991). Chicken forebrain Na + and Ca2+ channels and ligand-gated gama-aminobutyrate (type A anion channel), and quisqualate receptors have beenexpressed in Xenopus oocytes (Sigel and Baur 1988). Channel activities were22measured using voltage-clamp and intracellular recording techniques in thepresence of specific receptor agonists. Quisqualate perfusion over the cellsdecreased Na + (44%) and type A-gated (50%) currents while increasing currentthrough the Ca + channel (166%). This effect could be mimicked by the PKCactivators, 11-phorbol 12-myristate 13-acetate and the diacylglycerol analogue 1,2-oleoylacetylglycerol. Phorbol esters that do not stimulate PKC had no effect, andtamoxifen (a PKC inhibitor) blocked the response. To resolve whether or not thistype of channel modulation is an artifact of expression in an oocyte preparation,Numann et al. (1991) have measured the effects of phorbol ester anddiacylglycerol treatment on Na + currents in dissociated rat brain neurons. Usingwhole-cell voltage-clamp techniques they found that phorbol esters and/ordiacylglycerol analogue treatment could reduce the Na + conductance by 80%. Tobe certain that PKC was the effector of the Na + channels, inside-out membranepatches containing Na + channels were excised from the neuron. These patcheswere exposed to the same phorbol esters and diacylglycerol analogues as abovebut in the absence of PKC. This had no effect on the Na ± channel conductanceuntil purified PKC was added to the medium. This then resulted in a reduction ofNa + channel conductance which was similar to that observed in the whole-cell.Furthermore, the ATP dependence of this effect clearly establishes that thechannel is being phosphorylated by PKC, and provides strong evidence that PKCmodulates Na + channel conductance in rat brain neurons.Voltage-dependent K + channels have also been expressed in Xenopusoocytes (Busch et al. 1992). As above, when phorbol ester activators of PKCwere applied the K + current was inhibited (by 20 - 88% dependent on membranepotential) and when PKC activity was blocked with staurosporine channelinhibition was abolished.23Heart Na + and Ca 2+ channels have also been shown to be underphosphorylation control by protein kinases and by direct membrane delimited Gprotein interaction. Single-Na + channel current recordings from whole-cell patchclamps show a 35% reduction in current in response to the cAMP analogue 8-bromo-cAMP (Schubert et al. 1989). This response could also be elicited by theaddition of isoproterenol (a R-adrenergic agonist) to the cell bathing medium,suggesting that Na + channels are down-regulated by R-adrenergic stimulation.Furthermore, when purified G protein was added to inside-out excised patches theNa + channel current was also inhibited (Schubert et al. 1989). Calcium channelsin cardiac myocytes have been shown to be under a similar type of regulation.However, in this instance, isoproterenol treatment stimulates Ca 2+ channelcurrents. Treatment with phorbol esters could stimulate Ca 2+ currents andinhibitors of PKA could block this effect (Yatani et al. 1987). Interestingly, purifiedG protein could also stimulate Ca 2+ currents in inside-out excised patches.In addition to the adenylate cyclase - PKA mediated ion channel regulation,outlined above, a guanylate cyclase pathway has been identified and anothersignalling cascade elucidated via cyclic-guanosine monophosphates (cGMP) (Hille1992). Again these have been shown to have a role in ion channel regulation, inthis case of the nonselective cation channel of photoreceptors (Hille 1992).There exist then, indirect (via phosphorylation) and direct, pathways for theregulation of ion channel permeability by G proteins and cAMP and cGMPdependent protein kinases but the interplay between the many levels of regulationremains to be clarified.The plasma membrane Na + /K ATPase has also been demonstrated to beunder phosphorylation control. In renal proximal tubule cells (RPTC) Na + /KATPase is responsible for Na + resorption. When RPTC were incubated in thepresence of the protein kinase C activators, either synthetic diacylglycerol - L-a-l-24oleoyl-2-acetoyl-sn-3-glycerol or the phorbol ester - phorbol 12,13-dibutyrate, Na +resorption decreased by approximately 50%, measured in permeabilized cells bythe hydrolysis of 32P-ATP. A phorbol ester that does not stimulate PKC (4-a-phorbol didecanoate) did not inhibit Na ± /K + ATPase activity and sphingosine, aPKC inhibitor, abolished the inhibitory effect of the above PKC activators(Bertorello and Aperia 1989). Moreover, employing techniques similar thoseabove, Na + /K + ATPase activity in renal cortical collecting duct (RCCD) wasshown to be modulated by protein kinase A (Satoh et al. 1992). Forskolin anddibutyryl cAMP, which increase cAMP ; , inhibited Na + /K + ATPase activity byapproximately 60%. An inhibitory peptide (1P20), a specific inhibitor of PKA,abolished the inhibitory effects of forskolin and dibutyryl cAMP. Using purifiedshark rectal gland and partially purified rat renal cortex Na + /K + ATPasephosphorylation was examined by 32P incorporation (Bertorello et al. 1991). Bothprotein kinase A and C were found to phosphorylate the enzyme, and this wasabolished in the presence of specific inhibitors (PKI-(5-24)-amide, and Ca 2+removal from the medium, respectively). Phosphorylation by either of the kinaseswas also associated with a 40-50% decrease in the activity of the enzyme.This area of research is clearly in an early stage, and there is no clearsynthesis of all these regulatory interactions. However, in terms of mechanismsby which membrane ion permeability can be down-regulated this area offers themost promise. The anoxia tolerant hepatocyte preparation described here offersan excellent system in which these mechanisms can be studied, especially sinceso many specific ligands, inhibitors and stimulators of receptors, ion channels andprotein kinases are now known to exist (Hille 1992).25Anoxia Tolerant vs Anoxia Intolerant Liver Cell DeathIt is useful now to summarize the key events leading to cell death, andspecifically liver cell death since it is the tissue under investigation here, of ananoxia/hypoxia intolerant species. This will serves to illustrate the lesions, orevents that must be avoided by anoxia tolerant cells. In Figure 1 an outline of theevents leading to anoxic/hypoxic cell death in an anoxia/hypoxia intolerantmammalian liver is shown. With the loss of oxidatively derived ATP, and theresultant mismatch between ATP synthesis and utilization, cellular [ATP] rapidlydeclines. Due to the decline in [ATP], Na + /K + ATPase function is limited, thetranscellular Na + and K + gradients begin to dissipate, and the plasma membranebegins to depolarize. Then, what is thought to be the hallmark of cell death, anincrease in cytosolic [Ca 2+ ] occurs (Herman et al. 1990). The decreasingmembrane potential opens voltage dependent Ca 2+ channels, allowing Ca 2+ toflux into the cytosol. Calcium is also thought to leak from de-energizedmitochondria and endoplasmic reticular stores. The increased [Ca 2+ ]; activatesphospholipases which begin to degrade the plasma membrane phospholipids andfacilitates cell rupture and death. Whether or not a cytosolic increase of Ca2+ ,from either the extracellular space or intracellular stores (Herman et al. 1990) orboth, preceeds cell death is now in question. Evidence suggesting such a increasecomes from experimentally increasing [Ca 2+ ] ; and observing subsequent celldeath, or removing Ca 2+ from the cell or tissue bathing medium and delaying celldeath (see review by Herman et al. 1990) More recently, it has beendemonstrated that cytosolic [Ca 2+ ] does not rise, and that mitochondrialmembrane potential decreases without the release of Ca2+ , prior to cell death(Herman et al. 1990). The initiation of cell death in this instance wascharacterized by the appearance of blebs on the surface of the plasma membrane.26Figure 1. A summary of the probable sequence of events leading to cell death inan anoxia-intolerant mammalian liver and the supposed absence of such events inan anoxia-tolerant turtle liver. 1 Turtle liver [adenylate] are known to be maintainedfor at least 5 h of anoxia (Brooks and Storey 1988); however, the maintenance ofion gradients and decrease of ATP utilizing pathways can only be inferred. Inmammalian liver recent evidence indicates that membrane rupture occurs prior toan increase in cytosolic [Ca2- ] (Herman et al. 1990; dotted line 2). Alternatively,the solid line pathway indicates the sequence of events to cell death involvingincreased cytosolic [Ca 2+ ] from mitochondrial stores (dotted line 3) and from theextracellular space.2728Cell death was reversible up until the point when the blebs ruptured (Herman et al.1990). Since polymerization of microfilaments and microtubules requires theenergy of hydrolysis of ATP, it was suggested that the ATP deficiency, due to theloss of oxidative phosphorylation, resulted in cytoskeletal disruption and blebbing.Regardless of the precise sequence of events leading to cell death, in the turtleliver (and other tissues), this cascade of events must be avoided. Thus far it isknown that liver [adenylate] are maintained during anoxia (Brooks and Storey1988). The relationship between reduction in ATP turnover, decreased Na + /K +ATPase activity, and therefore ATP utilization, and the maintenance of iongradients is speculative and remains to be investigated.SummaryFrom the literature review above it is apparent that the current knowledge ofhypoxic defense mechanisms has been derived from a heterogenous assemblageof species. I reasoned that the complexity and diversity of these in vivo systemsprecludes detailed biochemical analysis and the comprehensive understanding ofanoxic defense strategies. A cell-level system from an anoxia tolerant speciessuch as the freshwater turtle C. picta would be invaluable. Accordingly, in thisthesis I set out to achieve the following main goals.(1) to develop and define conditions for the routine preparation of hepatocytesfrom the anoxia tolerant species Chrysemys picta bellii.(2) to define the magnitude and reversibility of the cellular metabolicsuppression induced by anoxia, in terms of ATP turnover.29(3) to define the response to anoxia of an ATP utilizing reaction, in particularthat of the Na + /K + ATPase, since a large proportion of the cellular ATPturnover is used to support ion pumping.(4) to define the response to anoxia of the cellular membrane potential.Within a cell system such as this the more universal intracellular signallingand defense mechanisms, like those underlying the metabolic arrest and channelarrest concepts, can be probed. Moreover, a comprehensive understanding of ananoxic defense strategy, which is more difficult to obtain from tissue or wholeanimal systems, can be obtained from a cellular system such as the hepatocytesystem presented here.30Chapter 2: Preparation and Characterization of Anoxia Tolerant Hepatocytes.PrefaceThis chapter is adapted from a paper published by L.T. Buck, S.C. Land and P.W.Hochachka (Am. J. Physiol. in press, 1993). A portion of the cell counting,cellular leakage of lactate dehydrogenase, cell protein measurements andmeasurement of cellular adenylates via HPLC protocols were performed with theassistance of S.C. Land. The majority of this work was done by myself, and allother aspects of data collection and presentation were performed by myself.IntroductionTo study strategies of anoxia tolerance a model is needed that is anoxiatolerant over long periods, homogeneous with regard to function, and wherecellular chemistry is readily measurable with conventional techniques. Rathepatocytes have been employed to study the metabolic alterations due to anoxia,but suffer from a problem common to all other systems employed. They are notanoxia tolerant, suffering a 90% decrease in [ATP] following 30 min of anoxia(Anderson et al. 1987, Aw and Jones 1989), and thus demonstrating the inabilityto regulate ATP utilizing and synthesizing processes in a coordinated fashionthroughout a normoxic - anoxic transition. Turtle heart and brain have beenextensively studied as anoxia tolerant models (Ultsch and Jackson 1982, Wasseret al. 1992, Lutz et al. 1985, Doll et al. 1991). From these studies it is apparentthat cellular ATP levels are preserved during anaerobic periods of up to 4 h, and acomparison of ATP production rates based on 02 consumption (aerobic) andlactate accumulation (anaerobic) data indicate that a coordinated reduction in31metabolism is occurring. However, these heart and brain preparations lack theconvenience of a cell level system in which mechanisms are more easily studied,and for this reason attention was focussed on the liver. Described here is amethod for the isolation of hepatocytes from the western painted turtle that aremetabolically competent, survive long periods of anoxia, and demonstrate acoordinated reduction in metabolic rate.Materials and MethodsHepatocyte Isolation. Female turtles (Chrysemys picta bellii) weighing between250-750g were obtained from WM. A. Lemberger Co., Inc. Oshkosh, WI andmaintained in aquaria at 15-20°C with a flow through dechlorinated water system.Animals were killed by a subcutaneous injection of sodium pentobarbital (200mg.kg -1 body weight). Following a 30 to 40 min period the plastron was quicklyremoved with an electric bone saw to expose two large abdominal veins. Thesevessels were quickly cannulated, and the liver was perfused at a rate of 3 ml.min -1at 20°C with a medium containing 1.4 mM Na2HPO4, 2.2 mM KCI, 78.5 mMNaCI, 10 mM Na HEPES and 34.5 mM NaHCO3 at pH 7.5 (Solution A). The initialperfusion continued for 25 min. Following cannulation the liver was dissectedfrom the animal and placed floating in a petri dish containing Solution A. The liverwas then perfused for 9 min with Solution B (Solution A with 2 Units.m1 -1 SigmaProtease XXIV). Digestive enzymes conventionally used in the preparation ofhepatocytes, collagenase, trypsin, and hyaluronidase were not effective indigesting turtle livers. Following perfusion the liver was minced with a razor bladein ice cold Solution C (Solution A with 2% bovine serum albumin, BSA), and theresulting slurry was poured successively through two nylon filters, a 253 pm filterand then a 73 pm filter. All centrifugations and subsequent steps were carried out32at 4°C. The cell suspension was centrifuged for 4 min at 50 x g, the resultingpellet resuspended in Solution C and centrifuged a second time for 2 min at 50 xg. The pellet was then suspended in Solution D (Solution A with 0.1 mM MgSO4,3.8 mM MgCI, 5.8 mM CaCl2 and 4% BSA) and centrifuged at 50 x g for 2 min.The resulting cells were suspended in Solution D, allowed to settle in a flatbottomed culture flask and stored at 4°C for a maximum of 24 h. Thecomposition of the liver perfusion media was based on previously measuredplasma ion concentrations in turtle (Jackson et al. 1984). From this size range ofturtles 1 g to 10 g (pelleted cell wet weight) of cells could be obtained. Liverweights were not usually taken in order that the cannulation be performed asquickly as possible, therefore yields were not rigorously calculated but estimatedto be 20% or less. All experiments reported here represent independenthepatocyte preparations.Cell Viability. Cell viability was tested 2 and 24 h after isolation. Hepatocytesstored for 24 h (as described above) were not resuspended in fresh media beforeviability tests. To assess trypan blue exclusion, 2 volumes of hepatocytes wereincubated for 5 min with 1 volume of 0.6% trypan blue and an aliquot counted forstaining in a Neubauer Chamber. Red blood cell and melanocyte contaminationwas also determined at this time and were always less than 1 %. Leakage oflactate dehydrogenase (LDH) was determined (as below) by centrifuging a 1 mlaliquot of cell suspension for 10 min at 100 x g and separating supernatant frompellet. The supernatant was assayed directly while the pellet was resuspended in50 mM imidazole, 2 mM EDTA (ethylenediamine-tetraacetic acid), 0.1% Triton pH7.0 and sonicated 2 x 10 sec each. Following centrifugation to remove cellfragments the supernatant was assayed for LDH activityAdenylate samples were obtained by injecting a 450 pl aliquot of cells into33an Eppendorf tube containing 50 pl of 70% perchloric acid maintained at -5°C inan ice-saltwater slurry. The resulting material was pelleted by centrifugation in arefrigerated microcentrifuge at 4°C for 10 min at 10,000 x g, the pellet wasdiscarded and the supernatant fraction neutralized with 1 M K2CO3. The resultingprecipitate was removed by centrifugation and the supernatant stored at -80°C fornot more than one week before being assayed. [ATP], [ADP], and [AMP] weremeasured using an LKB 2152 HPLC fitted with a 7pm Aquapore AX-300 weak ionexchange column (Brownlee Columns, Applied Biosystems, Inc. USA), aspreviously described (Schulte et al. 1992).Oxygen Consumption Measurement. Hepatocyte 02 consumption was measuredto obtain estimates of normoxic ATP turnover and recovery 02 consumption.Cells were taken from respective normoxic or anoxic incubations (see below). Analiquot of normoxic hepatocyte suspension (30 mg cells.m1 -1 ) was diluted 2 foldwith Solution D (normoxic) into a 2 ml thermostated (25°C) Gilson oxygen cell(Middleton, WI.) equipped with a Clark type 02 electrode (Yellow Springs, OH.).Recovery 02 consumption measurements were made from cells removed fromanoxic media by centrifugation (2 min at 50 x g, 22°C) and resuspended innormoxic media (approximately 30 mg cells.m1 -1 ). An aliquot was added to theoxycell as above.Enzyme Compartmentation. Tissue samples of approximately 0.5 g weredissected out and placed in 5 vols ice cold 250 mM sucrose, 2 mM EDTA and 20mM Hepes buffer at pH 7.2. The tissue was minced with the tip of a fine pair ofscissors and homogenized for 10 sec with an Ultra-Turrax homogenizer at amedium setting. The homogenate was centrifuged in a Sorval RC-5 centrifuge for10 min at 300 x g to remove large cellular debris (discarded). The supernatant34was then centrifuged for another 10 min at 8000 x g to pellet the mitochondria.The supernatant from the last centrifugation was decanted and assayedimmediately for phosphoenolpyruvate carboxykinase (PEPCK), citrate synthase(CS), and LDH activity. The pellet was resuspended in homogenization buffer andsonicated for 2 x 10 sec each, centrifuged and treated as above. All enzymemeasurements were performed using a Perkin-Elmer Lamba 3 Spectrophotometer(Norwalk, CT.) employing previously described methods (Suarez et al. 1985).The percentage of the enzyme activity (units.m1 -1 ; 1 unit =1 //mole substrateconverted to product.min -1 ) measured in each compartment was then applied tothe total activity measured to obtain the activity in units.g -1 in each compartment.The relative amounts of CS found in the cytosolic fraction and LDH found in thematrix fraction could be used to correct for the contamination of eithercompartment.Gluconeogenic Rate Determination. The rate of gluconeogenesis and oxidationfrom uniformly labelled 14C-lactate by isolated hepatocytes was measured bymethods previously described for salmonid hepatocytes (French et al. 1981).Freshly prepared hepatocytes were allowed to recover from isolation at roomtemperature for 2 h before gluconeogenic experiments began. Normoxichepatocyte incubations were maintained at 20°C with 10 mM lactate as substrateand 0.2 pCi of uniformly labelled 14C-lactate for 2 h. Labelled 14C-glucose wasseparated from other radioisotopically labelled compounds using Amberlite MB-3ion exchange resin (BDH Inc. Vancouver, BC.) as previously described (French etal. 1981) and counted with an LKB 1214 Rackbeta liquid scintillation counter. Todetermine if cellular signalling mechanisms were functioning, cells were incubatedas above with either 1 pM adrenalin, 10 pM glucagon, 1 pM insulin or 10 NMglucagon and 1 pM insulin.35Assessment of Anoxia Tolerance. Two 50 ml aliquots of Solution D werevigorously gassed, one with 95% air/5% CO2 (normoxic) and the other with 95%N2/5% CO2 (anoxic) for one hour prior to BSA addition. BSA (4%) was added toeach solution and the pH adjusted to 7.5 with NaOH under a flow of therespective gas mixture. Hepatocytes were removed from storage conditions andallowed to warm to 25°C and settle out of solution for one hour. Supernatantswere decanted and the cells resuspended in either of the above two solutions.Ten to 15 ml of cell suspension was then placed in 6 Erlenmeyer flasks (25 ml)with tight fitting rubber caps. A short piece of PE-100 tubing was insertedthrough a small hole in the cap, and a humidified anoxic or normoxic gas mixturewas perfused over the cells. A 25 gauge syringe needle was inserted through therubber cap to allow gas to escape. The 6 flasks were fixed to a stainless steelplate with spring clamps and the whole assembly placed into a plexi-glass waterbath. The incubating water volume was maintained at the same height as the cellsuspension in the flasks. This entire unit was placed on a orbiting shaker at 2cycles per sec. In this fashion triplicate anoxic and normoxic cell incubationswhere conducted over a 10 h time period at 25°C. After 5 h 0.5 mM potassiumcyanide was added to one anoxic and one normoxic incubation. At the same time10 mM sodium iodoacetate was added to another anoxic/normoxic pair. Also,serial 0.45 samples of cell suspension (cells and support medium) were taken at 0,2.5, 5, 7.5, and 10 h from each anoxic/normoxic pair to measure lactateconcentrations; serial 4 x 0.45 ml aliquots of cell suspension were also taken at 0,5, and 10 h for glucose, glycogen, adenylate, and p02 measurement. Metabolitesamples (lactate, glucose, and adenylates) were acid extracted and neutralized asdescribed above. Glycogen samples were acid extracted and centrifuged asabove. The supernatant and pellet were frozen in liquid N2 immediately and stored36at -80°C. Prior to assay, glycogen samples were thawed on ice and resuspended,and aliquots taken for amyloglucosidase digestion and background glucosemeasurement as previously outlined (Bergmeyer 1974). Samples for oxygenmeasurement were taken with a gas tight glass Hamilton syringe pre-rinsed withnitrogen bubbled Solution A and injected into a Radiometer D616 thermostated cell(25°C) (Bach-Simpson Ltd. London, ON.). The measured p02 from anoxicincubations was not different from Na-dithionite controls. Adenylates wereassayed using HPLC as described above. The other metabolite assays wereperformed using a Titertek Multiskan MCC/340 plate spectrophotometer (FlowLaboratories, Inc, Mississauga, ON.). All assays were adapted for use with theplate spectrophotometer from previously described methods (Bergmeyer 1974).Reagents were added to the wells and mixed using a Titertek Multichannel Pipette.Each microplate contained 96 wells of 0.3 ml each.Conversion Factors. For comparison to similar published experimental values: cellnumber.ml cell suspension -1 , cell number. mg wet wt cells -1 , and mg cellprotein.mg wet wt cells -1 were determined. The number of cells per volume ofsupport medium was determined by counting in a Neubauer Chamber. Wet cellweight was measured by centrifuging a known volume of cell suspension in apreweighed Eppendorf tube at 10,000 x g for 10 min, decanting the supernatantand removing any excess liquid with a Kimwipe tissue, and finally re-weighing thetube and pellet to determine the wet cell weight by difference. Hepatocyte proteinwas determined from an aliquot of acid extracted and neutralized stock cellsuspension (24) using a Bio-Rad Protein Assay kit (Mississauga, ON).Chemicals. Uniformly labelled 14C-lactate was obtained from Amersham CanadaLtd. (Oakville, ON.) or NEN Du Pont Canada Inc. (Mississauga, ON.). Radioisotope37was dried under N2 and dissolved in Solution D without BSA and usedimmediately. Amyloglucosidase and acetyl coenzyme A were obtained fromBoehringer Mannheim, Canada Inc. (Laval, Que.). Bovine insulin and glucagonwere obtained from Calbiochem (San Diego, CA.). All other chemicals werepurchased from Sigma Chemical Co. (St. Louis, MO.).Statistics. All statistics were performed using SYSTAT Version 5 (Evanston,IL.)software.ResultsAssessment of cell quality. Freshly isolated cells and cells incubated for 24 h at4°C were subjected to standard cell viability tests and the results are shown inTable 1. Under these conditions the cells never stained with trypan blue and LDHleakage accounted for only 1.7 to 3% of the total cellular activities. Importantly,the ATP content did not change, maintaining an energy charge (EC) greater than0.90 over the 24 h period. It was critical to determine if the hepatocytes weremetabolically competent; therefore, the rate of gluconeogenesis from lactate wasmeasured in normoxic hepatocytes (Table 2). Lactate was provided as thegluconeogenic substrate since it would be preferred over various amino acids onthe basis of the intramitochondrial compartmentation of PEPCK (Table 3)(Suarez etal. 1985). From initial 3 h time course studies the rate of gluconeogenesis wasfound to be linear after 1 h of incubation; therefore, all subsequent incubationswere carried out for 2 h. The rate of gluconeogenesis was 1.95 ± 0.04 pmolglucose.g cells -1 .0 and increased 80% by the inclusion of either 1 pM adrenalinor 10 pM glucagon, 3.48 ± 0.77 and 3.77 ± 0.93 pmol glucose.g cells -1 .0,respectively. Addition of 1 pM insulin did not significantly effect the38Table 1. Assessment of cell quality following isolation and 24 hours of incubationat 4°C (storage conditions) and in comparison to tissue values.Freshly Isolated 24 h@ 4°CTissueTrypan BlueStaining 0%(5) 0%(5)Enzyme Leakage l(LDH) 1.7 ± 0.07%(5) 3.0 ± 0.41%(4)Adenylates 2ATP 2.03 ± 0.10(4) 2.34 ± 0.06(4) 3.03±0.054ADP 0.21 ± 0.02(4) 0.26 ± 0.01(4) 2.00 ± 0.27AMP 0.069 ± 0.003(4) 0.017 ± 0.001(4) 0.75 ± 0.05EC3 0.92 0.94 0.70Values represent the mean ± the standard error and the number of independentcell preparations in parentheses.1 Percentage values are given as the percentage of the total cell number or totalcellular LDH activity.2 ymol adenylate.g cells -1 .3 Energy charge (EC) is defined as EC = [ATP] x 0.5[ADP]/[ATP] + [ADP] +[AMP].4 Values taken from Kelly and Storey (1988).39Table 2. The effect of adrenalin, glucagon, and insulin on the rate ofgluconeogenesis.Gluconeogenic Rate(pmol glucose.g cells -1 .0)Control 1.95 ± 0.04(5)Adrenalin (1 pM) 3.48 ± 0.77(5) 1Glucagon (10 pM) 3.77 ± 0.93(5) 1Insulin (1 ,uM) 2.81 ± 0.76(5)Glucagon (10 pM) +Insulin (1 pM) 3.35 ± 0.71(5) 11 Significantly different from control (One Sided Dunnett's Test, p <0.01), numbersin parentheses are the number of independent hepatocyte preparations used.Hepatocytes used in these experiments were freshly prepared.40Table 3. Total enzyme activities and the compartmentation of PEPCK and themarker enzymes CS and LDH in the liver of Chrysemys picta.Enzyme^Mitochondrial^ Total^Activity (%) 1^Activity2PEPCK 86 5.3 ± 0.62(8)CS 85 6.8 ± 1.28(8)LDH 4.8 95.0 ± 11.8(8)Note: values are the means ± standard error, with the numbers of animals used inparentheses.1 These values are reported as the percentage of total tissue enzyme activity.2Enzyme activities were measured at 20°C and are given as pmol.min -1 .g wetweight-1.41Table 4. Trypan blue exclusion, LDH leakage and energy charge after 10 hours ofanoxic or normoxic incubation (25°C) of hepatocytes in the presence or absenceof the inhibitors cyanide and iodoacetate.Normoxic AnoxicTrypan BlueExclusion %1 %Control 95.0 ± 2.3(4) 95.2 ± 2.1(4)Cyanide 90.3 ± 4.1(4) 84.2 ± 8.0(4)(0.5 mM)lodoacetate 1.2 ± 1.0(3)3 24.0 ± 20.0(3)3(10 mM)LDH Leakage %2Control 12.3 ± 0.4(3) 9.7 ± 2.4(3)Cyanide 10.0 ± 0.7(3) 10.4 ± 0.3(3)(0.5 mM)lodoacetate 59.3 ± 24.43(3)3 11.1^± 0.6(3)(10 mM)Energy ChargeControl 0.95 ± 0.01(4) 0.90 ± 0.01(4)Cyanide 0.95 ± 0.01(4) 0.90 ± 0.01(4)(0.5 mM)lodoacetate 0(3)3 0.18 ± 0.16(3) 3(10 mM)1 Percentage values are given as the fraction of cells excluding trypan blue relativeto the total number of cells. 2LDH % values are presented as the the activitymeasured in the cell suspension medium relative to the total activity (cellularactivity and LDH activity in suspension medium). Data points represent the mean± SE. with the number of independent hepatocyte preparations in parentheses.3Statistically different from respective control values (Tukey's, p < 0.05).42gluconeogenic rate or antagonize the glucagon stimulated rate. The rate of 14C-lactate oxidation was 1.32 ± 0.4 pmol CO2.g -1 .h -1 and can be only qualitativelyinterpreted due to possible randomization of label at several branchpoints (data notshown).Recovery Oxygen Consumption. Following a 10 h anoxic period hepatocytes thatwere resuspended in normoxic media had an 02 consumption rate that was notsignificantly different from control hepatocytes (Paired Student's t-test, p =0.05),0.19 ± 0.02(3) and 0.20 ± 0.02(6) pmol 02. g -1 .min -1 , respectively.Determination of Anoxia Tolerance. Adenylate concentrations (Fig. 2 A&B) weremeasured from cells incubated for 10 h under an atmosphere of either 95%Air/5% CO2 (normoxic) or 95% N2/5% CO2 (anoxic). There were no significantconcentration decreases in any of the adenylates over this time course undereither condition; moreover, the inclusion of 1 mM cyanide to a separateanoxic/normoxic pair did not alter the adenylate levels (Table 4). The cellular[ADP] did increase over the anoxic time course. The 5 and 10 h ADPmeasurements were significantly greater than the time 0 measurement (0.49 ±0.03, 0.49 ± 0.08, and 0.26 ± 0.01 pmol.g cells -1 , respectively). Cellular energycharge remained above 0.90 under control conditions and in the presence of 0.5mM cyanide. However, in the presence of 10 mM iodoacetate, EC dropped tozero under normoxic conditions and to 0.18 during anoxia (Table 4). Similarly,fewer cells excluded trypan blue under normoxia (1.2%) than anoxia (24%); andLDH leakage was greater under normoxia than anoxia, 59.3 and 11.1%,respectively. Trypan blue staining and LDH leakage in the presence of cyanidewere not significantly different from control values.43Anoxic Carbohydrate Metabolism. The rate of glucose and lactate production andglycogen depletion was measured over the same time course as above (Figs. 3, 4& 5). The rate of glucose production under anoxic conditions was 3 fold greaterthan under normoxic conditions 22.6 and 7.5 ymol.g cells -1 .0, respectively (Fig.3). The rate of lactate production under anoxic conditions (Fig. 4) was 4.2ymol. g -1 cells.h -1 , and this measurement gives an accurate measure of anaerobicglycolysis. The addition of 10 mM iodoacetate after 5 h of anoxic incubationcompletely blocked any further increase in [lactate] (data not shown). The rate ofglycogen depletion under anoxia was not found to be significantly different fromthe normoxic rate (Fig. 5).Conversion Factors. For comparison with similar published experimental valuesthe conversion factors were determined and are: cell number.ml cell suspension -1= 4.3 x 106 ± 4.6 x 10 5 (n =12); cell number.mg wet weight cells -1 = 1.3 x 106± 1.12 x 104 (n =4); and mg cell protein.mg wet weight cells -1 = 0.071 ± 0.01(n =7).DiscussionThe isolation procedure described results in a high yield of metabolicallycompetent cells, indicated by both direct and indirect tests. The direct tests showa constancy of adenylate concentrations during incubation, no increase in trypanblue staining, and less than 2% of the total LDH activity is in the extracellularmedium after isolation, all indicating that the cells are intact. The indirect testsshow that the cells effectively perform normal hepatocyte functions. They aregluconeogenic (this is a complex two compartment pathway), and respond to bothadrenaline and glucagon, confirming that cell surface receptors and intracellular44Figure 2. A comparison between normoxic (control 1A, 95% air/ 5% CO2) andanoxic (1B, 95% N2/5% CO2) adenylate concentrations and energy charge in 10hour control hepatocyte incubations. Data points represent mean ± SE. of 4independent hepatocyte preparations. Hepatocytes stored for no longer than 24hr prior to experiment. Values at 10 hrs are not significantly different from thevalues at time zero (Tukey's, p <0.05). Note that in some cases SE. bars fallwithin the size of the symbol.A. Normoxic0OcD4-0a) cr,0 —cOa)Q)0Ec-104 -3 -2 -0.8 a)rnca0.60.4 a)0.20.0086Vr0B. Anoxic•0--V-O 4-cD•47-,-4•••1CCD' s 3 -U —c T.DUv -c:7? 2• TDCOE^-cp  3 10-.00.8 a)rnco0.6 -c• (..)0.4 0)a)0.20.020 6 87 ,0 240 -T3 200 -a)^-0 160 -E^-120 -oco 80U^-40 -..30—46r I^,^I0^2^4^6^8^10Time (h)Figure 3. A comparison between glucose production from turtle hepatocytesincubated for 10 hours under normoxic (control, open circles) or anoxic (closedcircles) conditions. See figure legend 1 or materials & methods for conditions.Data points represent mean ± SE. of 6 independent hepatocyte preparations.50 -cnTu- 40 -c..)a)30 -E20 -a)co1 0 -co0I0^2^4^6^8^10Time (h)Figure 4. A comparison between lactate production in hepatocytes incubated for10 hours under normoxic (control, open circles) or anoxic (closed circles)conditions. See figure legend 1 or materials & methods for conditions. Datapoints represent mean ± SE. of 6 independent hepatocyte preparations.47— 900 -, co715 800 -0RI 700 -$7;E 600 -oc 500 -o)oc.) 400 -0 300 -1^1^1^I^i^I0^2^4^6 8 10Time (h)Figure 5. A comparison between cellular glycogen concentrations in hepatocytesincubated for 10 hr under normoxic (control, open circles) or anoxic (closedcircles) conditions, as in Fig. 1. See figure legend 1 or materials & methods forconditions Data points represent mean ± SE. of 6 independent hepatocytepreparations.4849signalling pathways are unimpaired. Together these properties attest to themetabolic integrity of the isolated cells. The obtained gluconeogenic rates weresimilar to rates reported for salmonid fishes (for overview see Suarez andMommsen 1985) but were, like fishes, lower than mammalian values (Elliot et al1976). The gluconeogenic rate in the presence or absence of glucagon wasunaffected by bovine insulin. This result is not completely unexpected since fishhepatocytes have demonstrated a similar insensitivity to bovine insulin (Foster andMoon 1987; Moon et al. 1985). As in the fish studies, reptilian hepatocytes arelikely to be more metabolically responsive to the species-specific insulinhomologue.The localization of PEPCK in liver tissue of a closely related turtle species(Pseudemys scripta) has been previously determined; however, in opposition to thevalue reported in Table 3 (86% mitochondrial) Penney and Kornecki (1973) reportthat 71 % of the PEPCK activity was cytosolic. This difference is easily explainablebased on the poor techniques employed in the earlier study. Briefly, tissue washomogenized for 5 min in distilled water, the homogenate was centrifuged for 60min at 24,500 x g, and PEPCK activity was assayed in both the supernatant andthe pellet. Firstly, the homogenization protocol would have compromised theintegrity of the mitochondria, resulting in leakage of mitochondrial enzymes.Secondly, without osmotic support the mitochondria may have ruptured, andthirdly no marker enzymes were measured; therefore, there was no estimate of thedegree to which mitochondria were damaged.Incorporation of radioisotope label into glycogen was not measured. It isreasonable to assume that a significant portion was not sequestered into glycogen,since evidence from fish and amphibian studies indicate that freshly isolatedhepatocytes are in a negative glycogen balance throughout incubation (Mommsenand Storey 1992; Moon et al. 1985) and even increase glycogenolysis in response50to adrenalin and bovine glucagon. Further evidence indicates that the amount oflabel that does become incorporated into the glycogen stores of fish hepatocytesis relatively small (Moon et al. 1985). Errors due to carbohydrate flux intoglycogen could lead to underestimating flux capacity of the gluconeogenicpathway, but these errors are likely small and do not alter the assumption that thebasic interpretation of the present gluconeogenic data is reasonable.Cells were typically prepared a day prior to use in an experiment, so it wascritical to assess the state of the cells after storage. Based on the criteria listed inTable 1 it is apparent that the cells were intact and storage conditions wereadequate. Cells were stored as long as 4 days without any changes in the aboveparameters (data not shown); however, they were never used for experimentsbeyond 24 h of storage.In comparison to rat hepatocytes, which when exposed to anoxia at 37°Close approximately 70% of their intracellular ATP in less than 10 min and only25% exclude trypan blue at 3 h (Andersson et al. 1987, Aw and Jones 1989),turtle hepatocytes survive 10 hours of anoxia at 25°C without any significant lossof adenylates or any significant increase in trypan blue staining or LDH leakage(Fig. 2, Table 4). The increase in [ADP] over the anoxic time course is likely theresult of insignificant changes in [ATP], since cellular [ATP] is 10 fold greater than[ADP]. The stability of the adenylate pool in the hepatocytes over this time courseis consistent with the findings of: Kelly and Storey (1988) for freeze clampedbrain, red/white muscle, liver and heart over a 5 h period; Wasser et al. (1990) for31 P-NMR measurement of ATP in the 3 h anoxic isolated perfused heart; and Lutzet al. (1985) for freeze clamped 2 h anoxic brain.Cyanide did not effect cell viability, as judged by ATP concentration, trypanblue staining or LDH leakage under either of the two experimental conditions(Table 4). In contrast, and consistent with brain, heart and whole animal51experiments (Perez-Pinzon et al. 1992, Doll et al. 1991, Lutz et al. 1985; Bing etal. 1972; Belkin 1962, respectively) iodoacetate, an inhibitor of the glycolyticenzyme, glyceraldehyde-3-phosphate dehydrogenase, caused [ATP] to decreasesharply, and trypan blue staining to increase significantly. Interestingly, anoxicincubation may have protected the hepatocytes to some extent from death byglycolytic inhibition. As shown in Table 4 the cells incubated under normoxiaexhibited the viability criterion of a population of dead cells, most notably theabsence of any measurable ATP. Alternatively, viability criteria from the anoxiccells indicate that the hepatocytes are maintaining some level of cellular integrity.These data also suggest the presence of mechanisms that protect these cells fromanoxic injury. For example, if as predicted from the channel-arrest hypothesis(Hochachka 1985), ion pumping and ion leaks were down-regulated, then theenergy demand of the cell would be reduced, slowing the onset of cell death andgiving the observed results.No exogenous substrates were added to the cell incubations; therefore, themetabolic inhibitor data isolate glycogen fermentation to lactate as the primaryenergy yielding metabolic pathway during the 10 h anoxic incubations. Cellularglycogen was depleted by approximately 25% over 10 h of anoxic incubation. Incontrast, rat hepatocytes exhaust intracellular glycogen after only 2 hours ofhypoxic exposure (Anundi and De Groot 1989). This is an important observationsince it has been previously observed that anoxic turtles can sustain theexhaustion of cardiac glycogen stores as long as liver stores are maintained (Dawet al. 1967). This indicates that the availability of liver glycogen is an importantfactor in determining anoxic survival time.The changes in lactate and glucose concentrations could account for thedecrease in glycogen over 10 h of anoxia. Although the glycogen depletion underanoxia was not significantly different from that of the normoxic rate, the increase52in [glucose] and [lactate] was. Therefore one must conclude that a differencedoes exit between anoxic and normoxic glycogen utilization. To be certain thatthe amyloglucosidase digestions of glycogen were complete, glycogen standardswere prepared and measured for glucose at 1, 2 and 4 h. Glycogen-standarddigests were complete by 2 h, and samples that were periodically remeasured after4 h of digestion were also complete by 2 h. The lack of any statistical significancecould be the result of measuring relatively small changes in the presence of a largebackground concentrations of glycogen (790 pmol glycosyl units.g cells -1 .A comparison of the rates of glucose and lactate production between theturtle and rat hepatocytes under anoxic conditions reveals that in the rat (Anundiand De Groot 1989), glucose production is 14 fold higher and lactate production18 fold higher, with 60% of the glycogen carbon recovered as glucose. In theturtle cells, however, 90% of the total carbon recovered is in the glucose pool.This suggests that the role of the liver as a substrate source during anoxia issimilar in the two species yet the ability to maintain cellular homeostasis duringanoxic periods is much greater in the turtle. Interestingly, blood glucose has beenshown to increase in Chrysemys picta during anoxia in response to catecholaminerelease (Keiver et al. 1992). Here, as with the fish, amphibian and rat hepatocytesdescribed above, catecholamines were absent yet the cells were exhibiting anegative glycogen balance. The means by which flux from glycogen to glucose isincreased is unknown but may be the result of a Ca2 + activated phosphorylase afollowing redistribution of intracellular Ca 2+ stores (Keppens et al. 1977). Therate of lactate production over the 10 h anoxic period agrees well with the valuesobtained from freeze clamped liver (Kelly and Storey 1988), assuming a linear rateof lactate production. The question of whether or not an early Pasteur effect ispresent in turtle tissues remains unresolved, since the rate of lactate productionwas not measured within a narrow enough time frame in this study, and the53forced submergence protocol of the above studies may result in a stress inducedshort-term rise in lactate concentrations.From the rate of lactate accumulation an anaerobic ATP turnover rate canbe calculated. Since there was no exogenous substrate present during anoxicincubations and the only endogenous source was glycogen, an ATP/lactate ratio of1.5 was used. This indicates an anoxic ATP turnover of 6.5 ± 0.43(6) pmolATP.g -1 .h -1 . From the measured 02 consumption and a P/0 ratio of 3 thenormoxic ATP turnover is calculated to be 68.4 ± 7.2(9) ,umol ATP.9 -1 .0(numbers in parentheses represent the number of independent hepatocytepreparations used). Based on these calculations the metabolic rate of these cellsin anoxia decreases by about 90%. This effect was completely reversible, forwhen anoxic cells were re-aerated 02 consumption rates similar to pre-anoxiclevels were obtained. Herbert and Jackson (1985a) calculated heat output for thewhole organism in an analogous fashion and obtained values at 20°C of 1.6 foraerobic and 0.28 cal.min -1 .kg -1 for anaerobic rates, representing a reduction inmetabolic rate (83%) similar to our value.A decrease of this magnitude is extraordinary and raises the obviousquestion. How is this event regulated and maintained? It is clear from the 90%decrease in ATP turnover rate (Table 5) that in order for cells to survive such longperiods (up to 10 h of anoxia at 25°C), there must exist a mechanism wherebyenergy utilizing and synthesizing pathways are down regulated in a coordinatedfashion. The sequence of events occurring in mammalian hypoxia sensitive cells(Fig. 1) following anoxic exposure are extensively described (Farber et al. 1981,Nayler 1983, Herman et al. 1990), and generally include: i) a precipitous decreasein cellular ATP levels, ii) plasma membrane blebbing, and iii) membrane rupture,loss of membrane ion gradients and subsequent cell death. The initial event (i.above) is the most commonly demonstrated; however, the sequence of events54that follow are not as well established and may not be universal for all cell types.From the available evidence (present paper, Kelly and Storey 1988), turtle livercells and other turtle cell types (heart, Wasser et al. 1991; brain, Lutz et al. 1985)do not appear to suffer a drop in [ATP] following anoxia and therefore should notbe subject to the above cascade of events. Thus, two distinct issues arise: firstlythe events of cell death or maintenance during short term hypoxic episodes (lessthan 30 min), and secondly a very different issue, regulated metabolic suppressionand survival over extended anoxic periods (in the case of the turtle, months). Theactual mechanisms of down-regulation of metabolism are complex and have beenrecently reviewed (Storey and Storey 1990). Briefly, there are three strategieswith supporting evidence regarding control of the energy synthesizing pathwaysoutlined in this review (and Chapter 1): 1) covalent modification of enzymes, 2)fructose-2,6-bisphosphate regulation of phosphofructose kinase (a rate limitingglycolytic step) and 3) reversible association of enzymes with particle-boundmultienzyme complexes. Also discussed are the mechanisms whereby energyutilizing pathways are also down-regulated and intracellular signals to initiate thecascade of events leading to a concerted down-regulation of metabolism;however, these mechanisms have less supporting evidence. The abovepossibilities have not been explored in anoxia tolerant vertebrates; therefore, theturtle hepatocyte preparation described here is well suited for this purpose.A provocative signalling mechanism elucidated from work on 02 limitedmammalian liver cells (which do not appear to down-regulate metabolism as much)is a decrease in intracellular pH. A hypoxia related decrease in pH and slowrecovery during normoxia, rather than a rapid recovery to normal pH values, offersprotection from degradative cellular processes and is presently the most plausiblehypothesis for anoxic survival for that system (for review see Lemasters et al.1987). However, Storey and Storey (1990) point out that a drop in pH is an55unlikely candidate for regulation in good animal anaerobes because these animalshave high tissue and blood buffering capacities and, where measured, intracellularpH decreases linearly over a period of hours to days, a process requiring muchmore time than the observed rapid depression of metabolism (Jackson 1968).In summary of Chapter 2, a physiologically competent anoxia tolerant cellpreparation for the study of metabolic regulation during reversible normoxic -anoxic transitions has been developed. This is the only hypoxia tolerant vertebratesystem at the cellular level of organization currently known.56Chapter 3: Microcalorimetric Measurement of Reversible Metabolic SuppressionInduced by Anoxia.PrefaceThis chapter is adapted from a paper by Buck, L.T., P.W. Hochachka, A.Schtin, and E. Gnaiger that is currently accepted pending modest revisions by Am.J. Physiol. I was trained in the use of a microcalorimeter by Dr. A. Schtin, whowas also involved as a "trouble shooter" in subsequent calorimetric experiments.Dr. E. Gnaiger provided a microcalorimeter and the laboratory space in which thecalorimetric experiments were conducted, as well as helpful advise. All otheraspects of data collection and manuscript preparation were performed by myself.IntroductionIt is generally assumed that in vertebrate tissues anoxic glycogenfermentation to lactate is the sole source of ATP for sustaining cell workfunctions. Since the caloric equivalent for lactate production via this pathway isknown (-77.7 kJ.mo1 -1 , accounting for the enthalpy of proton buffering; (Gnaiger1983)) and the rate of lactate production by anoxic hepatocytes previouslydetermined (Chapter 2), the amount of biochemical heat generated from thispathway can be calculated (-93 pW.g -1 cells). By using direct microcalorimetry itis possible to assess whether any additional heat releasing biochemical processeshave been overlooked (Curtian and Woledge 1978, Gnaiger 1980, Gnaiger andKemp 1990, Hardewig et al. 1991, Taetmeyer 1978).Microcalorimetry additionally provides an insight into the presence orabsence of a short-term Pasteur Effect. C. picta does not appear to undergo a57long-term Pasteur effect (Jackson 1968, Kelly and Storey 1988, Lutz et al. 1985)of the type observed in mammalian tissues; the expected 8 to 10 fold activation ofglycolysis is not observed (Hochachka 1987, Hochachka and Guppy 1987).However, without an estimate of aerobic glycolysis a definitive conclusion cannotbe made. There is some evidence of a short-term Pasteur Effect (Kelly and Storey1988) lasting less than an hour, which may represent a transition period to a newanoxic steady state. Uncertainty arises from whole-animal studies of this typebecause the elevated levels of lactate measured within this period may be stressinduced as a result of forced diving. Moreover, there may also be a tissue specificresponse to anoxia. Direct microcalorimetry should be capable of measuring anystepwise decreases in metabolism from the onset of anoxia and settle the issue ofa short-term Pasteur Effect.The aims of this Chapter were: (1) to measure the magnitude of the downregulation of metabolism calorimetrically to compare with biochemical data, (2) toassess the presence of any undetermined heat sources during anoxia (an anoxicgap), and (3) to investigate if anoxia-tolerant hepatocytes exhibit a short-termPasteur effect within the first hour or so of anoxia with cyanide.Materials and MethodsExperimental Animals. Turtles were obtained from W.M. Lemberger and Co. Inc.(Oshkosh Wi.) and maintained in fresh water (changed daily) aquaria at 15 - 20°C.Animals were killed by injection of sodium pentabarbitol (200 mg.kg  body weight -1 ) and hepatocytes were pre pared and stored by the method developed in Chapter2.Hepatocyte Preparation. Hepatocytes prepared on the same day or the nightbefore were allowed to settle out of solution and then resuspended in incubation58media (see Chapter 2) pre-bubbled with either 95% air/ 5% CO2 (normoxic) or95% argon/5% CO2 (anoxic) and allowed to equilibrate to 25°C. The incubationmedia consisted of: (in mM), 1.4 Na2HPO4, 2.2 KCI, 78.5 NaCI, 10 Na HEPES,34.5 NaHCO3, 0.1 MgSO4, 3.8 MgCI, 5.8 CaCl2 and, 2% BSA (wt/vol), pH 7.5.Microcalorimetry. The microcalorimeter used was a four channel Thermal ActivityMonitor (TAM, ThermoMetric AB, Sweden) equipped with two titration/perfusionstirring vessels (Suurkuusk and Wadsii 1982). Sample containers (3 ml vol.) werecharged with 2.7 ml of either anoxic or normoxic cell suspension (as in ref. SchOnand WadsO 1986) and lowered stepwise into the calorimeter as described byairman Nordmark et al. (1984). This procedure required 40 min and theequilibration time at the measuring position was about 30 min. The cells werecontinuously stirred at 80 - 100 RPM and gassed with the appropriate gas mixtureduring this time. When the chamber reached its final resting position the gassingwas terminated and the two stainless steel access cannulae sealed. For anoxicexperiments the calorimeter chamber was purged with the argon gas mixture (seeabove) for 5 min before the introduction of the stirring vessel assembly to preventpossible 02 contamination. Argon was chosen over nitrogen for theseexperiments since it is a heavier gas and would provide a better barrier against 02convection into the calorimeter chamber surrounding the sample vessel. Forexperiments where cyanide was introduced a stable recording was measured for 1- 2 hours before 0.5 mM NaCN was added via a cannula and electronicallycontrolled microinjection pump. Preceding normoxic, and following allexperiments, a 0.45 ml sample of cell suspension was taken and acid extracted foradenylate measurement as described in Chapter 2. All calorimetric experimentswere paired and ranged between 3 to 12 h of anoxia or normoxia. Typically,experiments were conducted using a hepatocyte density of about 20 - 30 and 4059to 60 mg cell(wet wt). m1 -1 , normoxic and anoxic experiments, respectively. Themeasurement of anoxic heat flux required a more dense hepatocyte suspension tomake it distinguishable from baseline.Static ampoules (5 ml) charged with 3 ml water were used as referencevessels. The instrument was calibrated electrically as described by GOrmanNordmark et al. (1984). For low temperature experiments the instrument was re-equilibrated to 10°C, allowed to stabilize for 24 h and re-calibrated.Oxygen Consumption Measurement. Hepatocyte 02 consumption was measuredto obtain an indirect measure of heat flux using the average oxycaloric equivalentof -450 kJ.mol 02, at 10 and 25°C. A measure of oxygen consumption was alsoused for the calculation of the ratio of calorimetric heat flux to respirometricoxygen flux (CR Ratio) for direct comparison to theoretical oxycaloric equivalents.Calculating a CR Ratio in this manner does not employ any assumptions aboutoxycaloric equivalents. For each hepatocyte preparation used in a calorimetricexperiment an aliquot of normoxic hepatocyte suspension was diluted 2 fold withincubation medium into a Cyclobios Oxygraph (Paar KG, Graz, Austria), totalvolume 4 ml, and thermostated at the appropriate temperature. Recovery 02consumption measurements were made from cells removed from anoxic media bycentrifugation (2 min at 50 x g) and resuspended in normoxic media (toapproximately 30 mg cells .m1 -1 ). An aliquot was added to the oxycell as above.Chemicals. All chemicals, including the enzyme for liver digestion (SigmaProtease XXIV), were obtained from Sigma Chemical Co. (ST. Louis, MO.).Statistics. Statistics were performed using an IBM compatible based program -Systat 5.0, Systat Inc. Evanston, IL.ResultsEnergy charge (EC = [ATP] + 1/2[ADP] / [ATP] + [ADP] + [AMP]) valuesranged between 0.74 and 0.87 (n =3 to 4) for the eight treatments: at 25°C and10°C, respectively for - normoxic, normoxic and cyanide, anoxic, anoxic andcyanide treatments. No significant differences were found amongst treatments(one-way ANOVA followed by Tukey's HSD, p =0.05). The addition of 2,4-dinitrophenol to normoxic and anoxic experiments did not decrease EC; however,iodoacetate addition resulted in the complete loss of ATP as described in Chapter2.Microcalorimetry. A typical calorimetric trace is shown in Figure 6 for aerobic andanoxic heat flow. The results of microcalorimetric measurements are summarizedin Table 5. There was a 76% reduction in heat flow when the cells were madeanoxic, a value that falls short of the 90% reduction in ATP turnover calculatedfrom biochemical measures (Chapter 2). The introduction of an anoxic cyanidesolution (27 pl to 2.7 ml) under aerobic conditions caused heat flow to decreaseby only 54%; heat flow under these conditions remained significantly higher thanwhen hepatocytes were made anoxic with N2 (Table 5). When cyanide was addedunder anoxic conditions, the resultant heat flow was not significantly differentfrom the pre-cyanide heat flow.To test the reversible nature of the metabolic suppression aliquots of stockcell preparations were re-aerated for one hour and placed in the oxygraph for 02consumption measurement. In the anoxic replicates, cells were typically heldanoxic for at least 2 h before re-aeration. Four anoxic batches of cells were re-aerated, and respiration was found not to be significantly different from aerobic6061Figure 6. A simultaneous calorimetric trace of normoxic (solid line) and anoxic(dotted line) hepatocytes at 25°C. Cyanide (0.5 mM) was added to eachincubation at the time indicated by the arrow. Anoxic cells are measured at twicethe concentration of normoxic cells to decrease the error at low heat flow, close tobaseline. Normoxic cells have a linear heat output over this concentration range.62Table 5. A comparison between direct and indirect measures of metabolicsuppression.Biochemical Determination^Calorimetric^Calorimetric(ymol ATP.g -1 .0)^(pW . M g 1 )^+ cyanide (0.5 mM)Normoxic 68.4 ± 7.2(9) 1 1.08 ± 0.08(8) 0.50 ± 0.04(4) 2Anoxic 6.5 ± 0.43(6) 1 0.26 ± 0.03(6) 0.23 ± 0.03(3) 21 Values from Chapter 2. 2Values significantly different from each other, (paired t-test, p <0.05).63cells (Paired Students t-test, p =0.05). 02 consumption of re-aerated cells was0.15 ± 0.02 and the control aerobic value was 0.18 ± 0.04 pmol 02.min -1 .g -1cells.In Figure 7 the 10 and 25°C data are presented on different axes, 7(A) toemphasize that the relative decreases due to anoxia and/or cyanide are similar atboth temperatures, and on the same axis 7(B), to show the absolute decrease inresponse to temperature and anoxia. The decrease due to temperature alonerepresents a Q10 of 4.2 over this range (calculated by the equation given bySchmidt-Nielsen (1984)) . A 15°C decrease in temperature results in an 88%decrease in heat flux, and the combined effect of a temperature reduction andanoxia decreases metabolism by 96%. The relative decrease due to an anoxicinsult is similar at each temperature, a 68% decrease at 10°C and 76% at 25°C.CR Ratio Determination. Aerobic heat changes calculated from 02 consumption at10 and 25°C were not significantly different from direct calorimetricmeasurements. The ratio of calorimetric heat flux and respirometric oxygen flux(CR-ratio) was -432 and -449 kJ.mo1 -1 , at 10 and 25°C, respectively (Table 6).This is compared to the theoretical oxycaloric equivalents for carbohydrate, fattyacids and amino acids spanning from -440 to -480 kJ.mo1 -1 02 (Gnaiger and Kemp1990). At an index of variation of the calorimetric and respirometricmeasurements of 7% (Table 6), the index of variation of the CR ratio is calculatedat 10% (Table 6). Therefore, with such large variation around the mean values adiscrimination between catabolic substrates of aerobic metabolism is not possibleon the basis of our calorimetric and respirometric measurements.Anoxic Heat Flux. When comparing heat flux under anoxic conditions with therate of lactate production, only 36% could be accounted for. This comparison is64Table 6. Oxygen consumption of hepatocytes at 10 and 25°C, and the ratio ofcalorimetric heat flux to respirometric oxygen flux (CR ratio).02 fluxnmol.mirfl .mg -102 fluxnmol.s -1 .g -1CR ratio 1kJ.mo1 -1 0225°C 0.15 ± 0.01(9) 2.5 -432 ± 43 210°C 0.026 ± 0.002(3) 0.43 -449 ± 77 2Values^are^the means^±^S.E.^with^the number of independent hepatocytepreparations in parentheses.1 Heat flux [mW. g -1 ] divided by oxygen flux [nmol.s -1 .g -1 J times 103 is the CRratio in units mJ././mo1 -1 or kJ.mo1 -1 .2 Not significantly different form theoretical oxycaloric equivalents, •kF102,calculated on the basis of various catabolic substrates (Gnaiger and Kemp, 1990)(paired t-test, p <0.05). The S.E of the CR ratio, SEC R , is calculated from theindex of variation (IV = SE/X) of the calorimetric (C) and respirometric (R)measurements,SECR = XCR ,/ivc2 + IVR265Figure 7. Normoxic and anoxic hepatocyte heat flux measured at 25 (n =9) and10°C (n =3) and as effected by 0.5 mM cyanide. Upper panel (A) shows heatoutput at the two temperatures plotted on different axes and panel (B) shows theheat output on the same axis.A.- 0.16 '7 w25°C^TD010°C 0.12E- 0.081._ 0.04 Lt..A- 0.00 CDa) • iaaaair•-1•••••Normoxic + KCN Anoxic + KCNCells^Cells0000066cn 1.21M,TD0 1.0E ■•=1B.U)a)ECUa)1.2 -1.0 - -25°C10°C••Normoxic + KCN Anoxic + KCNCells^Cells67based on the caloric equivalent of -77.7 kJ.mor l lactate, which includes the valueof -56 ILmo1 -1 lactate produced from glycogen (Gnaiger 1980) and the value of -21.7 kJ.mor 1 H + which is the enthalpy of neutralization of a proton by the HEPESbuffer (Beres and Sturtevant 1971). To investigate this discrepancy further,iodoacetate (10 mM) was added to anoxic cells as a test of the fraction of theheat flux that was due to glycolytic flux (Fig. 8). lodoacetate addition resulted in a25% reduction of heat flux, close to the 36% of the anoxic heat flux accountedfor by lactate production.An increase in heat flux in response to the addition of mitochondria!uncoupler would be indicative of mitochondrial metabolism under anoxia. In Fig. 9the results of the addition of 2,4-dinitrophenol, an oxidative phosphorylationuncoupler, are shown. Its addition to the anoxic hepatocytes resulted inapproximately a 30% increase in heat output.DiscussionBased upon the adenylate energy charge values obtained the hepatocytesremain viable during the normoxic and anoxic experiments. The hepatocytes showa large reduction in heat flux and therefore metabolic rate under anoxic conditions.If one assumes the response to cyanide mimicks the anoxic response it is a veryrapid effect also, taking less than 15 min. From the present calorimetric resultsthere does not appear to be a short-term Pasteur Effect as suggested by Kelly andStorey (1988) in this tissue, since there is a rapid and what appears to be anexponential decrease in heat output to a new steady-state that remains stable forat least 4 hours after cyanide addition. If a short-term Pasteur Effect was presentone would expect to observe a biphasic decline of heat flux, with a plateau in thecalorimetric signal at an intermediate level, corresponding to heat flux being68605040oLI302 20I100I I\/I\lodoacetate40^60^80^100^120^140^160Time (min)Figure 8. A calorimetric recording of anoxic hepatocytes at 25°C. lodoacetate(10 mM) was added where indicated by the arrow.69100.....-- 80560oit40cacyi20,,,,DNP•^ ■20 40 60 80 100 120 140 160 180 200Time (min)Figure 9. A simultaneous calorimetric trace of normoxic (solid line) and anoxic(dotted line) hepatocytes at 25°C. Where indicated 0.4 mM dinitrophenol (10 mMstock solution in H20) to and 0.5 mM potassium cyanide were added.070sustained by increased glycolytic rate. For this reason it is concluded that a short-term Pasteur Effect is not activated in turtle liver hepatocytes on normoxic-anoxictransition.To assess the presence of a long-term glycolytic activation a measure ofaerobic glycolysis is required. An estimate of glucosyl unit entry into glycoylsiscan be made from the rates of normoxic glucose release and of glycogen depletionreported in Chapter 2. The normoxic glycolytic rate is calculated to be 3.1 ± 4.5(5) pmol glucosyl unit. g -1 . h -1 and is not significantly different from the anoxicglycolytic rate based on the measurement of lactate accumulation, equivalent to2.2 ± 0.14 pmol glucosyl unit. g -1 . h -1 (paired t-test, p =0.05). The largestandard error in the above normoxic glycolytic rate value arises from thevariability in the measurement of cellular glycogen; therefore, biasing the statisticstowards a non significant result. However, if this is accepted as a reasonableestimate, the long-term response to anoxia does not appear to involve asubstantial Pasteur Effect. It is apparent though, that in no way does glycogenfermentation compensate for the loss of ATP from oxidative metabolism as isobserved in many hypoxia-sensitive tissues (Hochachka 1987).In response to anoxia hepatocyte heat flux decreased dramatically by about76%, a value that at both experimental temperatures, is close to the 85%decrease in heat flux measured by Jackson (1968) from whole-animalexperiments. Furthermore, the combined effect of a 15°C decrease intemperature (25 to 10°C) and anoxia resulted in a 96% drop in heat flux. Thisvalue approximates the whole animal reduction over a 20 - 3°C temperature rangeof 99.4% from oxygen consumption and lactate measurements (Herbert andJackson 1985). Normoxic Q 1 0 values calculated over similar temperature rangesalso agree, 4.2 and 3.8 (this study, and (Herbert and Jackson 1985),respectively). The anoxic Q 1 0 value from the above study determined from dived71turtles was 5.1, which is also close to the value of 3.5 presented here, especiallyconsidering the differences in protocol in these studies.When cyanide was added to an aerobic hepatocyte experiment the resultingnew steady-state heat flux was significantly different from the anoxic (argongassed) steady-state flux, with or without cyanide. However, cyanide addition didnot have a significant effect on anoxic cells. The anoxic heat flux of oxygendepleted (argon gassed) cells was an additional 50% lower than the aerobiccyanide inhibited cells. Cyanide specifically targets and effectively inhibits theactivity of cytochrome oxidase but does not inhibit peroxisomal 02 consumptionvia catalase, which may represent the additional heat output under theseconditions. Peroxisomal oxygen consumption has been estimated to beapproximately 10% of total hepatic respiration (Tolbert and Essner 1981), and thisobservation could account for the difference between the two measurements.Unfortunately, there are no accurate estimations of the heat generated by thispathway, but the enthalpy change per unit of oxygen consumed can be expectedto be close to the oxycaloric equivalent of respiration. Furthermore, when cyanideadditions were made to hepatocytes respiring in an oxygraph, 10-20% of the pre-cyanide oxygen consumption remained (data not shown).Under aerobic conditions indirect calorimetry approximates direct calorimetryclosely (Gnaiger 1983), and no significant difference was observed between thetwo measurements at either temperature in the isolated hepatocytes (Table 5).The isolated cells are different in this respect from most transformed cells inculture which show a significant anaerobic contribution to total heat flux underfully aerobic conditions (Gnaiger and Kemp 1990, Sch6n and Wads6 1986). Inthese mammalian cells, the combined enthalpy changes calculated from oxygenconsumption and lactate production balance with the measured total heat change(Gnaiger and Kemp 1990). However, from the comparison of the measured heat72flux of anoxic hepatocytes with the estimation of the heat flux from lactateproduction and the associated proton buffering, a large discrepancy was observed.When the calorimetric value was compared with the measured rate of lactateaccumulation (Chapter 2), only 36% of the anoxic heat could be accounted for.Previously, Gnaiger (1980) found that less than 50% of the calorimetricallymeasured heat flux could be accounted for by measuring anoxic end products.The discrepancy could not be explained by the heat change due to trace amountsof residual oxygen, yet heat flux and oxygen flux were in complete agreementunder aerobic conditions (Gnaiger and Staudigl 1987). Shick et al. (1988)confirmed the finding of a significant "exothermic gap" in heat flux under anoxia,concluding that there are yet unknown anoxic heat sources in the intertidal musselMytilus edu/is. In contrast, such a large exothermic gap was not observed in themarine worm Sipunculus nudus (Hardewig et al. 1991), where 83% of the anoxicheat flux could be accounted for.In the present study, one possible unaccounted for heat source could be thereversal of succinate dehydrogenase to function as a fumarate reductase aspreviously described in anoxic invertebrates and vertebrates (Hochachka et al.1975, Holwerda and de Zwaan 1979, Taetmeyer 1978). However, the possibilityof a major portion of the missing heat flux being the result of succinateaccumulation can be dismissed since it has been shown that its concentration inliver, heart and blood of 28 day 5°C anoxically dived turtles increased to only 2%of that of the lactate (Buck and Hochachka 1993).One of the main functions of the turtle liver under anoxic conditions is tosupply substrate, in the form of glucose from glycogen breakdown, to othertissues (Chapter 2, Clark and Miller 1973, Daw et al. 1967, Keiver et al. 1992).This pathway must contribute to some degree to the overall anoxic heat output.In Chapter 2 the rate at which anoxic hepatocytes release glucose was also73determined (17.1 pmol glucose.g cells -1 .0, (Chapter 2)). Despite the large Gibbsenergy change associated with the hydrolysis of glycogen to glucose, thecorresponding enthalpy of reaction is small, -3.10 kJ.mo1 -1 a,11-D glucose (aq)formed from animal glycogen (Gnaiger 1980). Using the rate of glucose releaseabove, therefore, the contribution of this pathway is calculated to be -0.015 /./W.g-1 cells, that is 6% of the heat dissipation measured under anoxia, but 16% of theenthalpy change calculated from the rate of glycoylsis to lactate and theassociated enthalpy of neutralization. Taken separately, the enthalpy ofneutralization amounts to 10% of the total anoxic heat change.The proportion of the glycolytic contribution (lactate formation) to anoxicmetabolism is further supported by the experiment shown in Fig. 8. The additionof iodoacetate to anoxic hepatocytes causes a further 25% decrease in heat flux,reinforcing the above interpretation.As shown in Fig. 9, DNP addition caused heat production to increase underboth normoxia and anoxia. Under normoxia this is an expected result and is dueto the increased oxygen flux generally observed upon uncoupling oxidativephosphorylation. The increase under anoxic conditions is suggestive of themaintenance of a mitochondrial pH gradient. Even though succinate was shownto accumulate to low levels relative to lactate, it is possible that in the presence ofuncoupler this pathway is activated (not explored in this study). Furthermore,Scott and Nicholls (1980) have described the reversal of the mitochondrial HATPase in mammalian synaptosomes supported by glycolytically derived ATP as apossible mechanism to generate an electrochemical gradient.In summary, hepatocytes from the aquatic turtle undergo a rapid andreversible metabolic suppression in response to anoxia. No evidence of a short-term Pasteur Effect was found and due to the rapid exponential decay in heat fluxthere does not appear to be one within the first 15 min following the onset of74anoxia (with cyanide). Furthermore, from the comparison of normoxic and anoxicglycolytic rates there is no apparent long-term sustained glycolytic activation. Thediscrepancy between calorimetric and biochemical measures of anoxic energy fluxcannot be attributed, in this system, to experimental artifacts due to residualoxygen uptake, as tested by the addition of cyanide under anoxia. Therefore, it isconcluded that during anoxia there are undescribed heat sources. From a previousstudy it is known that the fermentation of glycogen to succinate does notcomprise a significant portion of anoxic metabolism (Buck and Hochachka 1993);however, due to the increase of anoxic heat production in response to DNP somefacet of mitochondrial metabolism remains operative in anoxia.75Chapter 4: Suppression of Na + /K + ATPase Activity and a Constant PlasmaMembrane Potential in Hepatocytes during Anoxia: Evidence in Support of theChannel Arrest Hypothesis.PrefaceThis chapter is adapted from a paper by L.T. Buck and P.W. Hochachka currentlyaccepted pending modest revision by Am. J. Physiol. All aspects of datacollection and presentation were performed by myself.IntroductionThe western painted turtle (Chrysemys picta be//ii) is widely used as a modelsystem, in which the responses to, and the mechanisms of, anoxia tolerance arestudied (Chapter 1). One goal of this research has been to characterize andunderstand the biochemical processes involved in the coordinated down-regulationof metabolism associated with this species' response to anoxia. Thus far theability to greatly reduce metabolism and survive long anoxic periods has been wellcharacterized, and more recently the mechanisms by which this is achieved arebeing elucidated. Recent evidence points to regulation of membrane proteinchannel densities as one facet of the energy conserving mechanism (Hochachka1986, Pertz-Pinzon et al. 1992). Indirect and direct evidence for a functionalreduction in the membrane density and/or activity of Na t , K t , and Ca + channelshas been obtained (Bickler 1992, Lutz et al. 1985, Përdz-Pinzon et al. 1992a,POrez-Pinzon et al. 1992b). However, it is important to distinguish betweenmembrane protein channel activities and densities. For example if membranechannels were removed from the plasma membrane by endocytosis (as are muscle76plasma membrane glucose transporters (Goodyear et al. 1991)), a change indensity likely would not be detected employing radioisotopic ligand bindingmethods, and a potentially important reduction in ion influx could be overlooked.On the other hand, if the transport activity of a channel were measured, importantinformation about functional changes in ion flux would be obtained, but noregulatory information would be gained.It was, therefore, of interest to determine if there were any functionalchanges in the activity of the plasma membrane Na + /K + ATPase in response toanoxic incubation, since it is primarily responsible for the large electrochemicalgradient across the plasma membrane and its response to anoxia has notpreviously been determined. It was reasoned that if pathways for ion flux into thecell decrease in density and/or activity in anoxia one might expect that thepathway generating the ion gradient would be down-regulated coordinately withenergy supply (Chapter 1, Chapter 2, Land et al. 1993) and with channel densities(Përëz-Pinzon et al. 1992a) or activities. It has been previously shown thatneuronal membrane potential in turtle brain remains relatively constant in responseto anoxia or cyanide administration (Doll et al. 1991, Perez-Pinzon et al. 1992b)and that plasma membrane ion permeability decreases (Bickler, 1992, Lutz et al.1985, Per&-Pinzon et al. 1992a, Pèróz-Pinzon et al. 1992b). This would implythat Na + /K + ATPase activity also decreases, since the capacity fortransmembrane ion flux decreases and membrane potential remains constant, inresponse to anoxia. However, a direct comparison between Na + /K + pumpingunder normoxic and anoxic conditions has not been made. The absence of thispiece of evidence is likely due to the difficulty of measuring Na + /K + ATPaseactivity in intact tissues and tissue slices; in contrast the system developed here,an isolated hepatocyte preparation (Chapter 1), is well suited for this purpose.77In Chapter 2 it was shown that total normoxic ATP turnover rates decreaseby approximately 90% in turtle hepatocytes in response to anoxia; furthermore,protein and urea synthesis also are found to decrease by approximately 92 and72%, respectively (Land et al. 1993). Protein and urea synthesis were calculatedto comprise between 15 and 30% of the total anoxic ATP turnover. Ion pumping(via Na + /K + ATPase) is known to require a large fraction of the total cellularenergy supply, and it was therefore of interest to determine what fraction of theremaining anoxic ATP turnover could be accounted for by this pump. In otherwords, how important is the maintenance of ion pumping, and therefore iongradients, during anoxia?The plasma membrane Na + /K + pump is responsible for establishing acellular membrane potential, and it was of interest to know if changes in itsactivity would be translated to changes in membrane potential of anoxichepatocytes. Moreover, if membrane potential were unchanging and Na ± /K +pump activity decreasing, this would provide insights into the possibility offunctional channel arrest. Therefore, in addition to measuring pump activity,plasma membrane potential was also measured.The aims of the research in this Chapter are: 1) to directly measure theactivity of the Na + /K + ATPase in response to anoxia, 2) to assess the fraction ofATP turnover utilized by this pump during normoxia compared to anoxia, and 3) tomeasure plasma membrane potential in response to anoxia.Materials and MethodsAnimals and hepatocyte preparation. Spring and summer female turtles(Chrysemys picta bell) weighing between 250-750 g were obtained from WM. A.78Lemberger Co., Inc. Oshkosh, WI and maintained in aquaria at 15-20°C with aflow-through dechlorinated fresh water system.Hepatocytes were prepared the day before an experiment as described inChapter 2. Freshly isolated cells were suspended in sterile filtered support medium(pre-gassed with 95% air/5% CO2), placed in a culture flask, and stored at 4°Cuntil needed. The hepatocyte suspension medium consisted of (in mM): 1.4Na2HPO4 , 2.2 KCI, 78.5 NaCI, 10 Na HEPES, 34.5 NaHCO3, 0.1 MgSO4, 3.8MgCI, 5.8 CaCl2, and 2% BSA at pH 7.5. Prior to an experiment cells wereremoved from low temperature storage and allowed to warm to 25°C while beinggently shaken on a metabolic shaker for 2 h. Hepatocytes were then subdividedinto two aliquots and allowed to settle out of solution. Once settled thesupernatants were decanted and the cells resuspended in support medium (asabove) that was either: A) anoxic, pre-gassed with 95% N2/5% CO2 or B)normoxic, pre-gassed with 95% air/5% CO2. Hepatocytes were routinely adjustedto approximately 30 mg cells.m1 -1 support medium. For all experiments cells wereincubated at 25°C in capped 50 ml Erlenmeyer flasks with a constant flow ofeither humidified anoxic or normoxic gas mixture over the cell suspension. Cellswere prevented from settling out of suspension by gentle shaking on a metabolicshaker. As a measure of cell viability a 0.45 ml sample of cell suspension wastaken and acid extracted before and at the termination of each normoxic andanoxic experiment and assayed for ATP, ADP and AMP as described in Chapter 2.Adenylate concentrations are expressed as a ratio: energy charge (EC) = (ATP +Y2ADP) / (ATP + ADP + AMP).Measurement of Na + /K+ ATPase Activity. Hepatocytes were incubated for 3 has described above either anoxically or normoxically, except that the twoincubations were further subdivided into replicate normoxic and anoxic79incubations. This resulted in a total of 4 separate incubations of 9.5 ml of cellsuspension each in 25 ml Erlenmeyer flasks, 2 anoxic and 2 normoxic (gassed asabove). Three hour incubations were chosen because the previous calorimetricstudies (Chapter 3) indicated that anoxic cell suspension heat flux was in steadystate from approximately 15 min to 8 h. The initiation of each normoxic andanoxic experiment was staggered alternately by 1 h to allow for sampling. Twentyminutes before the 3 h mark ouabain was added to one anoxic and one normoxicincubation at a final concentration of 1 mM. In this way the net uptake of 86Rb +can be determined, from the difference between the total 86Rb 4- uptake (ouabainabsent) and the ouabain inhibited uptake (nonspecific uptake), as described below.A substock ouabain solution was prepared fresh each time by dissolving the stockpowdered ouabain into 1 ml of anoxic support medium (without BSA) and heatingon a hot plate. The solution was not allowed to boil and the final concentration ofouabain was 40 mM. It was cooled to 25°C before use.At the 3 h mark 0.5 ml of anoxic 86Rb + substock solution (specific activity3.0 x 106 CPM.pmol Rb +-1 ) was injected through the caps of two hepatocyteincubations to a final activity of 2-3 pCi.m1 -1 and one hour later to the other set oftwo flasks. This resulted in a final 86Rb + concentration of 43 pM, and insubsequent calculations the dilution of "RI) + by K ± is accounted for.Immediately after the final injection was made 1 ml aliquots of cell suspensionwere obtained and vacuum filtered (250 mm Hg) and washed with 1 ml of eithersolution A or B through 8 pm cellulose filters (Millipore Ltd., Mississauga, ON.) toremove the supernatant. A filter manifold and collection box (Hoefer ScientificInstr. San Francisco, CA.) were used to collect the cells. Samples were obtainedat 0, 5, 10, 20, 30, 40, and 50 min intervals. After vacuum filtering for 4 min(when the filter and cells appeared dry) the cells and paper were placed intoscintillation vials containing 10 ml water and 1% Triton X-100. The vials were80counted by Cerenkov radiation counting in a LKB 1214 Rackbeta liquid scintillationcounter. Sample quench was not significant and the counting efficiency was99.9%.To control for the non-specific binding of 86Rb + to filter paper and cellularfragments, an aliquot of the substock 86Rb + solution and an aliquot of sonicatedcells that had been pre-incubated in the presence of 86Rb + as above were vacuumfiltered. Filter papers retained less than 1% of the substock 86Rb + activity, whilesonicated cells retained 10-20% of the activity found in non-sonicated intact cells.Each curve shown in Figure 10 represents the difference between cell populationsincubated in the presence and absence of ouabain; therefore, a net uptake of86Rb + was determined, and non-specific binding automatically corrected for bythe difference calculation.Membrane Potential Measurement. To measure membrane potential via thedistribution of 36CI - across the plasma membrane it is necessary to obtain ameasure of intra and extracellular volumes. Therefore, four 25 ml Erlenmeyerflasks containing 4 ml of hepatocyte suspension each were incubated for 3 hunder either anoxic or normoxic conditions as outlined above; one anoxic/normoxicpair for cell volume measurements and another pair for membrane potentialmeasurements. Intra and extracellular volumes were measured using 1 pCi.m1 -13H20 and 0.5 pCi.m1 -1 14C-polyethylene glycol (PEG), respectively, as previouslydescribed (Rottenberg 1979) Membrane potential was measured from thetransmembrane passive distribution of 36CI - and 3H20 (Nobes and Brand 1989).An initial membrane potential time course was taken to determine if the potentialwas changing with time; measurements were made at 10, 20 40, and 60 minutes.Cells were separated from supernatant by oil centrifugation. One ml of cellsuspension was layered on top of 300 pl of oil mixture (dimethyl phthalate : di-iso-81octyl phthalate, 1.1 : 5) in a 1.5 ml Eppendorf tube and centrifuged at 12000 x gfor 30 sec at room temperature (22°C). After centrifugation 100,u1 of supernatantwas taken and placed in a scintillation vial containing 10 ml of ACS II scintillationfluid for counting. The remaining supernatant, oil and pellet were immediatelyfrozen in liquid N2. After approximately 30 min tubes were removed, and the tip,containing the cell pellet, was chopped off with a razor blade and hammer. Thecut tip was immediately placed into 10 ml of ACS II scintillation fluid and sonicatedat the highest setting for 20 sec. Radioactivity in supernatants and pellets wasdetermined by dual-channel liquid scintillation counting for 3H and 14C, and 3 H and36CI by using quench and crossover corrections where appropriate. Countingefficiencies for 3H, 14C, and 36CI were approximately: 42, 92, 99%, respectively.The oil layer was determined to retain less than 5% of the total radioactivity;therefore, the small amount that remained with the pellet in some instances wasdisregarded.Chemicals. Solutions of 86RbCI and ASC II scintillation cocktail were obtainedfrom Amersham Canada Ltd. (Oakville, ON) and Na 36Cl, 3H20 and 14C-PEG (as adry powder) were purchased from Du Pont Co. Inc. (Wilmington, DE). Dimethylphthalate and di-iso-octyl phthalate oils were obtained from BDH Chemical Inc.(Poole, Eng.). All other chemicals were obtained from Sigma Chemical Co. (St.Louis. MO).ResultsAs a measure of cellular integrity adenylates were routinely measured afteranoxic and normoxic incubations. Adenylate energy charge was not significantlydifferent (Paired t-test, p =0.05) between anoxic and normoxic 3h incubations,820.85 ± 0.03(6) and 0.87 ± 0.03(7), respectively. These results are consistentwith the previous long-term anoxic turtle hepatocyte incubations in Chapters 2 and3, and Land et al. 1993, anoxic turtle brain studies (Lutz et al. 1985), and isolatedanoxically perfused turtle heart (Wasser et al. 1990).Ouabain lnhibitable 86Rb + Uptake. The net 86Rb + uptake into normoxic andanoxic hepatocyte suspensions was determined, and the data are presented inFigure 10. The net uptake of 86Rb + was used to estimate the activity of theNa 4- /K + ATPase and was calculated from the slope of the curves in Figure 10.The respective slopes of the normoxic and anoxic curves were 37.9 ± 6.0 and9.6 ± 2.7 cpm 86 F _ +.o mg cells -1 .min -1 . Based on the standard stoichiometry of 3Na + transported out for 2 10 - into the cell and the concomitant hydrolysis of 1ATP, the Na + /K + ATPase specific ATP turnover was calculated under bothexperimental conditions. The ATP turnover values were calculated to be 19.1 ±4.8 and 4.8 ± 0.8 pmol ATP.g cells -1 .0, normoxic and anoxic cells, respectively(Table 7). The proportion of total cellular ATP turnover that is due to Na + /K +ATPase activity was determined and is shown in Table 7. The normoxic andanoxic total ATP turnover values have been previously determined (Chapter 2).Under normoxic conditions Na + /K + ATPase activity comprises 28% of the cellulartotal ATP turnover, while under anoxia it comprises 74% (ouabain inhibitable 02consumption was 31.3 ± 5.2(4)%; 02 consumption was 0.19 ± 0.03(4) pmol02.g-1 min-1‘ .) Furthermore, when the depression of cellular ATP turnover iscompared to the decrease in demand by the Na + /K ± ATPase, the total turnoverdecreases to a greater extent, thus giving rise to the greater proportion of totalcellular ATP turnover by the ATPase under anoxia.From the above data it is clear that in anoxia the ability of liver cells togenerate an ion gradient via Na + /K ± pumping has decreased to one quarter of8312800 -~y 2400 -T)c.) 2000 -cy)E 1600 -+cc-Q 1200 -Ea 800 -U40000^10^20^30^40^50Time (min)Figure 10. Net uptake of 86Rb + by anoxic (open circles) and normoxic (closedcircles) hepatocyte suspensions over a 50 min time course. Hepatocytes werepre-incubated for 3 h under either normoxic or anoxic conditions (25°C) before86Fr- +0 (2-3 pCi.m1 -1 ) was added to the incubations. The results shown representthe ouabain inhibitable 86Rb + uptake from 4 (anoxic) and 6 (normoxic) pairedindependent hepatocyte preparations.84Table 7. Proportion of Total Cellular ATP Turnover Utilized by the Na + /K +ATPase under normoxic and anoxic conditions.Normoxic Anoxic ReductionTotal 68.4 ± 7.2(9) 1,2 6.5 ± 0.4(6) 2 90%Na + /K + ATPase 19.1^± 4.8(6) 4.8 ± 0.8(4) 75%Proportion 28% 74%1 Units are pmoles ATP. g cells -1 . h -1 . Values represent the mean ± SE with thenumber of independent hepatocytes preparations in parentheses. 2 Values fromChapter 2. Slope values used to calculate Na + /K + ATPase activity obtained froma regression analysis through the entire data set comprising Fig. 10, not throughthe mean values shown therein.85normoxic levels. The questions arise: does plasma membrane potential decrease,or does it remain constant? If the latter occurs, then by inference ion influxpathways must be functionally down-regulated (channel arrest, Hochachka 1986)as in turtle brain (Pertz-Pinzon et al. 1992). To investigate this possibility plasmamembrane potential was measured in both normoxic and anoxic hepatocytesuspensions.Membrane Potential Measurement. A time course for 36CI - equilibration acrossthe plasma membrane was determined and is shown in Figure 11. Over the 10 to60 min time course 36CI - distribution and therefore membrane potential remainedconstant at approximately -27 mV. Following the 3 h normoxic or anoxic pre-incubations all subsequent 36 CI - incubations were continued for 20 min. Table 8shows the results of the normoxic and anoxic experiments. There were nosignificant differences between normoxic (-31.3 ± 5.8 mV) and anoxic (-30.6 ±3.9 mV) membrane potentials. Furthermore, when 0.5 mM cyanide was included,to prevent any re-oxygenation artifacts during transfer to tubes for oilcentrifugation, there was no significant difference between this value (-28.3 ± 3.2mV) and the above two values.To depolarize the cell and to be certain plasma membrane potentials werebeing determined accurately, normoxic hepatocyte suspensions were incubated inthe presence of a 150 mM K + medium and valinomycin (Table 8). Valinomycin isa K + ionophore and will cause cellular depolarization in the presence of 150 mMextracellular [K + 1. The membrane potential was not significantly affected by 150mM [K 41 medium alone, but when combined with valinomycin, membranepotential decreased to approximately one half of normoxic control or anoxicvalues.5E-40 --35 --30 --25 --20 --15 --10 --5 - 860^0 — 1^I^I^I^1^11 0 20^30^40^50^60Time (min)Figure 11. A constant plasma membrane potential determined by 36CI - distributionacross the plasma membrane of normoxic hepatocytes in suspension.Hepatocytes were pre-incubated for 3 h under normoxic conditions at 25°C beforethe addition of 36CI - and 3H20. The results have been corrected for 36CI - in theextracellular pellet space as described in materials and methods. The results wereobtained from 4 independent hepatocyte preparations.87Table 8. A comparison of hepatocyte plasma membrane potential under anoxicand normoxic conditions and as effected by high K + and valinomycin medium.Membrane Potential(mV)Normoxic -31.3 ± 5.8(4) 1Anoxic -30.6 ± 3.9(4)Anoxic + CN -28.3 ± 3.2(4)High K + (150 mM) -27.3 ±^1.2(5)High K + + Val (100 pM) -15.6 ± 1.7(6) 21 Numbers in parentheses are the number of independent hepatocyte preparationsused. 2Significantly different from normoxic control, p =0.02; all others notdifferent from normoxic control, p<0.05.DiscussionMetabolic suppression in response to anoxia has been previously observedin C. picta (Edwards et al. 1989, Jackson et al. 1984, Ultsch and Jackson 1982)and hepatocytes isolated from this species (Chapter 2, 3). It has also been shownthat protein synthesis, degradation, and urea synthesis all decrease to nearimmeasurable levels during anoxia (Land et al. 1993); therefore, it was of interestto investigate remaining energy utilizing pathways. Energy budget studies ofhepatocyte metabolism show that the plasma membrane Na + /K + ATPase utilizes alarge fraction of the total cellular energy demand (6-40%, (Nobes et al. 1989)).One would deduce that the activity of this pump must decrease in response toanoxia, since it has been demonstrated that ATP concentration remains constantwhile ATP synthesis rates decrease by 90%. Therefore, the proportion of cellularanoxic ATP turnover used for Na + /K + pumping was determined.There has been some criticism of the methodology employed to estimateNa + /K + ATPase activity from ouabain inhibited oxygen consumptionmeasurements (Nobes et al. 1989). Hepatocyte preparations in general probablygive the most consistent and accurate estimates of liver Na + /K + ATPase becauseoxygen, substrate, and inhibitor diffusional limitations, as well as more complexsampling protocols inherent in tissue and tissue slice preparations, are overcomeby working in cell suspensions. A serious consideration is depletion of cellularNa + and K + gradients in response to long-term ouabain incubations as suggestedby Nobes et al. (1989), which results in decreased respiration from Na + co-transported substrates (such as fatty acids). Nobes et al. (1989) point out thatouabain inhibition is very rapid (less than 1 min) at 37°C in rat hepatocytes, andlong-term ouabain preincubations are unnecessary. In fact, it can give rise to over8889estimates when oxygen consumption, in the presence of a Na + co-transportedsubstrate, is used to determine the fraction of total ATP turnover comprised ofNa + /K + ATPase activity. The present inhibitor experiments involved a 20 minpreincubation with ouabain because it was difficult to measure a net Rb + uptakebefore this time (which could be due to the overall low metabolic rate ofectotherms and specifically reptiles, (Else and Hulbert 1987)). Na + /K ± ATPaseactivity was measured directly from Rb + uptake in the absence of exogenoussubstrate; therefore, substrate co-transport could not diminish the Na ± gradientand result in substrate limitation and decreased oxygen consumption. In addition,Na + /K + ATPase activity has been previously measured in liver slices from a lizard(Amphibolurus vitticeps) at 37°C (Else and Hulbert 1987) and was found tocompromise 35% of the total cellular ATP turnover, close to the value of 28%measured here.Remarkably, Na + /K + ATPase activity in hepatocytes from C. pictaundergoes a 75% decrease in response to anoxic incubation (Table 7). Atpresent, it is not known how a decrease in Na + pumping of this magnitude isachieved, although an interesting possibility that could apply here has beenelucidated from isolated kidney tubule preparations (Satoh et al 1992) and in vitroNa + /K + ATPase studies (Bertorello et al. 1991, Tria et al. 1974). These studieshave shown that the pump can be inhibited by 40 to 60% by phosphorylation ofthe a subunit. Phosphorylation (and thus inhibition) is mediated by PKC and canbe prevented by sphingosine (a PKC inhibitor, (Bertorello et al 1989)). Furthersupporting evidence comes from the ability of phorbol esters, which are specificPKC activators (Bertorello et al 1989), to induce pump inhibition. Protein kinase Ahas also been implicated in phosphorylation and regulation of Na + /K + ATPase(see Chapter 1). Allosteric modification of key glycolytic enzymes, and down-regulation of their activity, has been previously put forth and demonstrated to90occur in marine molluscs and turtles as a general mechanism of down-regulation ofmetabolism (Brooks and Storey 1989, for review see Storey and Storey 1990). Ingeneral, these pathways are thought to be initiated by changes in intracellular[Ca2+ ] and can be activated through cell surface receptors. However, theincrease in cellular [Ca 2+ ] necessary to facilitate this sequence is not observed inneurons from turtle brain slices (Bickler 1992). Turtle hepatocytes in suspensionclearly demonstrate a metabolic suppression (Chapters 2, 3) in response to anoxiaand a down-regulation of Na + /K ± ATPase activity in the complete absence ofexogenous mediators. Therefore, an intracellular signalling mechanism must exist;however, these mechanisms are currently unknown.The energetic cost (in terms of ATP equivalents utilized) of maintaining iongradients by the hepatocytes is reduced to one quarter of normoxic levels inresponse to anoxia. However, without a parallel reduction in ion influx pathwaysthe cell would rapidly depolarize, and this would not be an expected consequenceof an anoxia tolerant system. Functional changes in ion influx can be assessedindirectly by measuring plasma membrane potential under both normoxic andanoxic conditions. When the plasma membrane potential was measured, therewas no statistical difference between normoxically or anoxically incubated cells.The value obtained (about -30 mV) is similar to values previously obtained from rathepatocytes (microelectrode technique, -32 mV, (Fitz et al. 1989); 36CI-distribution, -27 mV, (Nobes and Brand 1989)) and toadfish hepatocytes (S 14CN -distribution, -33.8 mV, Walsh 1989)). Liver plasma membrane contains a C1 -/HCO3 - transporter, and there is a possibility that it interferes with the accuratemeasurement of plasma membrane potential using 36CI - distribution. Nobes andBrand (1989) have compared the 36CI - distribution with the equilibrium potential ofK + , where plasma membrane potential was set by varying extracellular [K 1 ] in91the presence valinomycin, and found that the 36CI - distribution was an accuratemeasure of membrane potential.Since membrane potential did not change in response to anoxia (alsoreported for turtle neurons (Doll et al. 1991, Pêrtz-Pinzon et al. 1992►), eithermembrane ion channels or Na + /K + pumps must be functionally down-regulated.This could be achieved by either a decrease in activity or density (or both) of ionchannels or Na + /K + pumps. Recently, a 42% decrease in Na + channel densitieshas been demonstrated through the use of 3 [H]brevetoxin (a Na + channel specificligand) in turtle cerebellum in response to anoxia (Pèrez-Pinzon et al. 1992).Furthermore, there is indirect evidence demonstrating a decrease in Ca 2+ (Bickler,1992) and K + (Lutz et al. 1985, Perez-Pinzon et al. 1992) leakage from anoxicturtle neurons. Moreover, mammalian neuronal ion channels have also beenshown to be under allosteric regulation by cAMP mediated pathways (Sigel andBaur 1988) and by the direct and indirect interaction with G proteins (Brown1991). The work of Perez-Pinzon et al. (1992) clearly demonstrates a decrease inNa + channel densities; therefore, in this instance the possibility exists thatcoordinated down-regulation of the Na + /K + ATPase is achieved by the decreasedsodium influx. This is a possibility because pump activity is very sensitive tointracellular sodium concentrations (Cohen and Lechene 1990, Rossier 1987).Moreover, Edwards et al. (1989) using a turtle brain synaptosome preparationdemonstrated that 02 consumption could be increased or decreased by increasingNa + channel activity with veratridine or closing them with tetrodotoxin,respectively. This, implies regulation of the Na + pump via intracellular [Na +] andconcomitant modulation of cellular 02 consumption. From the present results it isunclear as to whether the Na ± pump is under allosteric control, is removed fromthe plasma membrane by endocytosis or simply lacks stimulation via Na + re-entry.Chih et al. (1989) raise the possibility that Na + /K + ATPase activity could also92decrease in response to lower ATP concentrations in the microenvironment of thepump.A final and important point to emphasize is that it is now possible toaccount for most of the ATP turnover in the anoxic turtle hepatocyte: Na + /K 4-ATPase activity accounts for about 74%, while and protein and urea synthesisaccounts for about 15-30% of the total ATP turnover (Land et al. 1993) underthese conditions. The energy required for cytoskeletal maintenance, Ca 2+ ATPaseactivity, RNA synthesis, and other functions remains undetermined, butpresumably must add up to only 10% or less of the anoxic ATP turnover rate.In summary of Chapter 4, the Na 4- /K + ATPase activity decreases by 74% inturtle hepatocytes in response to anoxia. Under normoxia the ion pumping activityof the Na + /K + ATPase requires 28% of the total ATP turnover of the cell and inresponse to anoxia it is reduced but now contributes to a much larger fraction ofthe anoxic ATP turnover (75%). This implies that ion pumping, and thereforemaintenance of ion gradients is important for surviving anoxia. Furthermore,plasma membrane potential does not change significantly in response to anoxia, aresult which suggests that ion channels are functionally down-regulated in somemanner, as demonstrated for Na + channels in anoxic turtle cerebellum (Perez-Pinzon et al. 1992). It still remains unclear if the actual number of functionalNa ± /K + ATPase copies are actively suppressed or if the pump activity is simplyreduced in response to a decrease in the Na + influx and intracellular Na ±concentrations. Importantly, the down-regulation of an energy utilizing reaction(Na + /K + ATPase) in a coordinate fashion with energy production has beendemonstrated.93Chapter 5. General Discussion.A detailed discussion follows each research chapter.^Therefore, thefollowing is a summary of important research findings and a discussion of how acoordinated reduction of metabolism might be achieved.Research SummaryA method was developed for the isolation of anoxia tolerant hepatocytes(primary culture) from the western painted turtle (Chrysemys picta bell). Thecells undergo a reversible 10 fold decrease in ATP turnover without a concomitantchange in ATP concentrations (Chapter 2). Over a 10 h period of anoxicincubation at 25 °C there were no significant changes in hepatocyte [ATP] or cellviability parameters (trypan blue staining and LDH leakage) unless iodoacetate wasadded at the 5 h mark. The addition of a glycolytic inhibitor caused [ATP] todecrease and a change in the viability parameters that indicated cell death.lodoacetate inhibition and subsequent cell death, in the absence of exogenoussubstrate, also indicates that glycogen fermentation is the major anaerobic ATPproducing pathway during anoxia. Moreover, the accumulation of lactate to highlevels in the hepatocytes and suspension medium suggests that lactate is themajor anaerobic product . Using direct calorimetry it was of interest to verify themetabolic suppression observed from anoxic and normoxic ATP turnovercalculations. At 25°C heat flux decreased by 75% in response to anoxic pre-incubation (Chapter 3). This is different from the 90% decrease calculated fromATP turnover in response to anoxic incubation. Only 42% of the exothermic heatflux could be accounted for from lactate accumulation and glucose release; theremaining 58% was from an unknown source, giving rise to an exothermic gap.94The lack of a biphasic decrease in heat flux in response to anoxia (imposedby cyanide administration) suggests there is no short-term Pasteur Effect in thefirst 20 min of anoxia. Subsequent calculations of aerobic and anaerobic carbonflux through glycolysis indicate that there is no long-term sustained glycolyticactivation either (Chapter 3).It is critical to long-term anoxic survival that rates of ATP utilization bedown-regulated in concert with rates of ATP synthesis. It was shown that ATPsynthesis rates decrease in response to anoxia (above), and in Chapter 4 ionpumping by the Na + /K 4" ATPase was shown to decrease by 75% in response toanoxic incubation. Since ion pumping was measured directly, by estimating theouabain inhibitable fraction of 86Rb 4- uptake, and since the distribution of manyions across the cell membrane ultimately depends on Na + /K + ATPase, theseresults demonstrate a functional decrease in the activity of an ATP utilizing cellfunction in response to anoxia. Of particular interest is the observation thatapproximately 90% of the anaerobic ATP turnover can be accounted for and thatapproximately 75% of this can be attributed to the Na + /K + ATPase (15%representing the anoxic component of protein synthesis and urea synthesis). Thisindicates that ion pumping is a larger percentage of overall ATP-consumingprocesses during anoxia than in normoxia. Most importantly, these data suggestthat there is a coordinated down-regulation of metabolism; that is, decreasingoxygen availability signals a reduction in ATP production and utilization. Since ionpumping markedly decreased, and Na + /K ± ATPase activity is responsible forestablishing a membrane potential, it was of interest to measure membranepotential in anoxic hepatocytes. If the normoxic membrane potential wasmaintained during anoxia, while Na ± /K + ATPase activity decreased, an acutedecrease in plasma membrane ion permeability would be suggested. When plasmamembrane potential was measured, by the transcellular distribution of 36Cr, there95was no difference between the two experimental conditions. This evidencesupports the notion that in anoxia tolerant species, such as the freshwater turtle,not only are cell membranes less permeable to ions in comparison to anoxiaintolerant species, but that there is a mechanism for further decreasing membranepermeability to ions in response . to anoxia or energy limiting conditions - i.e.,channel arrest.In the discussion that follows: 1) the coordinate regulation of ATP producingpathways with ATP utilizing pathways will be re-examined in the context ofcurrent kinetic models and concepts of metabolic regulation, and 2) based onempirical data a schematic diagram of regulatory pathways in a hepatocyte ispresented that integrates intra and extracellular signals with metabolic responses(Figure 11). A schematic diagram such as this will also provide direction for futureresearch, such as determining the functional importance of: 1) enzymephosphorylation/ dephosphorylation, 2) enzyme binding to the cellular particulatefraction, and 3) intracellular signalling mechanisms; all of which are not easilyassessed in more integrated tissue and whole-animal systems.1) Regulation of a Coordinated Suppression of Energy MetabolismAs discussed above, to sustain a long-term down-regulation or up-regulationof cellular energy metabolism, ATP production and utilization must be balanced.Typically, to understand changes in rate, enzymes and entire pathways have beenfitted to one of two well known kinds of kinetic models: 1) Michaelis - Mentenkinetic models, in which, for any change in rate there is a proportional change inthe concentration of substrate, intermediate metabolite or product, giving ahyperbolic velocity (pmol substrate.min -1 ) vs substrate concentration (v/ES1) curve.2) Allosteric models in which site-site positive cooperativity leads to sigmoidalv/[S] curves which can be further modified by enzyme-specific metabolic96modulators. These models have been used to aid in the understanding of the 10-100 fold increase in ATP turnover observed in mammalian skeletal muscle during arest-work transition. However, the first model can only accommodate modestactivity changes since the proportional change in substrate metabolites that wouldbe expected when large changes in flux occurs are not observed (Hochachka andMatheson 1992). Whereas the second model can account for greater activitychanges (because the allosteric regulator can increase or decrease enzymeturnover in a way that is not simply 1:1 with [substrate]), there are no knownallosteric mechanisms that can account for the 100 fold increase at all enzymesteps in ATP turnover observed during intense work in mammalian skeletal muscle(Hochachka and Matheson 1992). An alternative hypothesis suggests that inresponse to increased energy demand by working muscle, effective enzymeconcentration increases. This would result in higher metabolite flux rates withoutrequisite changes in metabolite concentrations. However, no such mechanisms,with one or two exceptions, for the regulation of effective enzyme concentrationare currently known (Hochachka and Matheson 1992).It is also difficult to explain the extreme metabolic depression observed inturtle liver by a kinetic model. The main changes in concentrations of glycolyticmetabolites in anoxia turtle liver are noted for glucose-6-phosphate (G6P),pyruvate, and lactate (Kelly and Storey 1988). All three of these metabolitesincrease by approximately 2 fold. The magnitude of this change is not sufficientto explain the observed 10 fold decrease in ATP turnover observed in turtlehepatocytes (Chapter 2). Furthermore, G6P is also an intermediate to free glucoserelease from the liver, and an increase in [G6P] and [glucose] probably reflects aincrease in the activity of the glycogen to glucose pathway. The observedincrease in [pyruvate] is likely a mass action effect at the lactate dehydrogenaselocus due to the increase in [lactate] and [H ± ], both anoxic endproducts. Finally,97liver concentrations of fructose-2,6-bisphosphate, a potent activator ofphosphofructokinase, do not change significantly in response to 5 h of anoxia(Brooks and Storey 1989). Taken together, the lack of proportional changes inmetabolites to ATP turnover and the lack of a decrease in the known potentactivator of PFK, indicate that the 10 fold decrease in ATP turnover cannot befitted easily or readily to either of the kinetic models, although there clearly is needfor more data here.Covalent modification of key glycolytic enzymes has been shown to alterenzyme kinetics in a way suggestive of a less active enzyme form (Storey andStorey 1990). However, phosphorylation of turtle liver PK in response to anoxiahad no significant kinetic effects, but phosphorylation of PFK from this tissueresulted in a 1.5 fold increase in its K m for F-6-P (Brooks and Storey 1989). Thisis accompanied by a 40% increase in [F-6-P] (substrate) which essentiallyeliminates any kinetic advantage covalent modification may have conferred (Kellyand Storey 1988). Thus, the functional significance of PFK phosphorylation inresponse to anoxia remains to be determined. It is interesting to postulate thatcovalent modification of PFK and PK alters the temperature sensitivity of theseenzymes. This could result in higher Q 1 0 values and lower metabolic rates overthe naturally occurring low winter temperatures this species encounters.The lack of proportional changes in intermediate metabolite concentrationsand significant changes in the concentration of the PFK stimulator F-2,6-P in theearlier studies indicate that the observed anoxia induced 10 fold decrease in ATPturnover cannot be explained on the basis of currently popular kinetic models.This however, is consistent with the lack of a glycolytic flux change observed here(Chapter 3) and suggests that it is the rate of ATP utilization that is more potentlyregulated in response to anoxia.98Singly the most significant change in energy metabolism during thetransition to anoxia is the loss of oxidative phosphorylation. An intermittentchange in [ATP] and [ADP] would almost certainly be expected; however, theexponential decrease in heat flux (Chapter 3) and the long-term stability of theadenylate pool (Chapter 2; Land et al 1993) in response to anoxia indicate thatlarge changes in [ATP] and [ADP] are not occurring. Although there was nosignificant decrease in [ATP], there was a significant 2 fold increase in [ADP](Chapter 2; Land et al. 1993). Since [ATP] is 10 fold greater than [ADP], astatistically non-significant change in [ATP] may result in significant changes in[ADP]. Why a glycolytic activation is not observed, evidenced by calorimetry andlactate accumulation, in response to a 2 fold increase in [ADP] is unclear.It is clear however, that to balance the 10 fold reduction in glycolytic ATPproduction a corresponding decrease in ATP utilization must necessarily occur.Therefore, it is essential also to consider the mechanisms by which ATP utilizationcan be down-regulated in response to oxygen limiting conditions. A possibilitypredominant in the literature is the temporary breakdown of ATP to the level ofadenosine. Adenosine is thought to bind to extracellular adenosine receptors toinitiate an intracellular response, possibly involving phosphorylation of glycolyticenyzmes and membrane ion channels and pumps. This is a provocativemechanism because in isolated turtle cerebellum, brain slices, and hepatocytepreparations there is a decrease in ATP turnover in the absence of exogenouseffectors. In support of the adenosine mediated mechanism, transient andstatistically significant changes in [ATP] and [ADP] have been measured in turtlebrain heart, red muscle, white muscle, and liver within the first hour of anoxia(Brooks and Storey 1988: Lutz et al. 1984). Subsequently, [ATP] and [ADP]recover to their pre-anoxic state or in some instances to higher concentrations.More recently, extracerebral adenosine levels after 100 min of anoxia have been99shown to increase 20 fold to 20 auM (Nilsson and Lutz 1992). Moreover, whenaminophylline (an adenosine receptor antagonist) was administered to anoxiccrucian carp the rate of ethanol production increased 3 fold, suggesting thatadenosine is involved in metabolic suppression in this species. From ischemicmammalian brain studies two lines of investigation point to the release ofadenosine as an important compound in ameliorating anoxic brain damage. First,the adenosine analogues cyclohexyladenosine, R-phenylisopropladenine, and 2-chloroadenosine reduced ischemic brain damage; and secondly, the A-receptorantagonists theophylline, caffeine, and dipropylcyclopentylxanthine increaseischemic brain damage as evidenced by LDH release, [Ca 2+ ]; increase, and therapid loss of ATP and phosphocreatine (PCR) (Rudolphi et al. 1992). Adenosinetherefore, is a promising candidate for an endogenous effector that mediates thecoordinated down-regulation of metabolism.The major goals of this thesis were: (1) To develop and define conditions forthe routine preparation of hepatocytes from the anoxia tolerant species Chrysemyspicta bellii, (2) To define the magnitude and reversibility of the cellular metabolicsuppression in terms of ATP turnover, (3) To define the response to anoxia of anATP utilizing reaction, in particular that of the Na + /K + ATPase, and (4) To definethe response to anoxia of the cellular membrane potential. As discussed above,these have been largely achieved, and we have a cell level system in which ATPturnover decreases 10 fold in response to anoxia and importantly ATP utilizationby the Na + /K + ATPase also decreases, by 4 fold. Furthermore, ion pumpingdecreases by 75% in the absence of a change in membrane potential (Chapter 4).These data require mechanisms that regulate membrane functions, such as ionchannel densities or permeabilities and ion pumping, in response to anoxia. Henceit is tempting to end my discussion by piecing together a testable model of the100turtle hepatocyte that may account for the response observed on normoxic -anoxic transition.2)^Pathways Regulating Coordinated Metabolic SuppressionIn Chapters 1 and 4 specific examples from the literature of reversiblephosphorylation of plasma membrane ion channels and Na + /K + ATPase as well asmembrane delimited interactions of G proteins with ion channels are reviewed.Presently, these types of intracellular second messenger mediated reponses can beused to formulate a model or framework for a universal and rapid mechanism forthe coordinate down-regulation of ATP producing and utilizing pathways.According to this model one would predict that anoxia tolerant species such as C.picta have an enhanced capacity to: 1) covalently modify ion channels and pumpsand/or 2) activate membrane delimited inhibitory G proteins in order to facilitate areduction in membrane coupled ATP utilizing functions and/or 3) covalently modifyand down-regulate key glycolytic enzymes.In Figure 12 a schematic diagram is shown that outlines known pathways ofsecond messenger mediated enzyme phosphorylation and membrane delimited Gprotein interactions with ion channels. Shown are transmembrane receptors, theirassociated second messengers, and protein kinases; not shown (in order tosimplify the diagram) is adenylate cyclase. The second messengers shown inFigure 12 have been documented in liver (Hepler and Gilman 1992; Cohen 1992).In the ever burgeoning field of receptors, second messengers, and G proteinsmany receptor subtypes have been indentified (Rudolphi et al. 1992; Exton 1985;Hepler and Gilman 1992) and many more are likely to be described (Yeun andGarbers 1992). However, only two receptor subtypes are considered here, thoseof the adenosine receptor - A1 and A2, since the apparent absence of a liver A1-receptor (Shimizu 1983; Londos et al. 1980) could be important to the101Figure 12. A schematic diagram of the possible metabolic control pathways andregulatory sites involved in the coordinated down-regulation of energy metabolismin a turtle liver cell. Shown are the plasma membrane receptors: a adrenergic, 1adrenergic, and adenosine (A2); the intracellular 2nd messengers and associatedprotein kinases (PKA and PKC); inositol triphosphate (IP3), diacylglycerol (DAG),and protein kinase C (PKC); cyclic adenosine monophosphate (cAMP), and proteinkinase A (PKA►; not shown is adenylate cyclase which couples receptors via Gproteins to the formation of cyclic monophosphate. Activation of the variousprotein kinases can effect cellular [Ca2+ ], the phorphorylation state of glycolyticproteins [glycogen phosphorylase (GP), phosphofructose kinase (PFK), andpyruvate kinase (PK)1; the phosphorylation state of ion channels and Na + /K +ATPase (phosphorylation denoted by -P). Also shown is the membrane delimitedinteraction of a G protein subunit with a transmembrane receptor segment(representing all above receptors which have been shown to interact directly withG proteins) and the subsequent interaction with an ion channel. Solid linesindicate known interactions, dotted lines indicate less well known interactions.The binding of glycolytic proteins to the cellular particulate fraction is not shown,but the functional role of binding interactions may be important to regulation toenergy metabolism. Nomenclature of G protein, 2nd messengers, and proteinkinases as used by Cohen (1992) and Billman (1992).102103construction of a concerted down-regulation mechanism. The A l -receptor hasbeen implicated in the activation of inhibitory G proteins, which have the effect ofdecreasing ion flux through a non-specific cation channel and inhibiting IP3formation. The A2-receptor, on the other hand, has been shown to activateadenylate cyclase leading to the stimulation of PKA and activation ofphospholipase C with the concomitant formation of IP3, DAG and stimulation ofPKC (Rudolphi et al. 1992). As mentioned in Chapter 1 PKC and PKA have beenshown to phosphorylate and modulate the activity of ion channels and pumps,respectively.Although stimulation of adrenergic a and fl receptors activate proteinkinases that subsequently affect the activity of enzymes, ion channels, and ionpumps, in an isolated hepatocyte preparation such as the one described here (andother isolated turtle tissue studies) receptor agonists are absent from theincubation. In vivo however, circulating catecholamines are important to thewhole-body anoxic response, since catecholamine levels have been shown toincrease by 10 - 20 fold during an anoxic dive, and the rate of glucose releasefrom the liver was significantly reduced by the f3 receptor antagonist propranolol.Phentolamine, an a receptor antagonist, had no effect (Keiver and Hochachka1992).Presently then, there are two possibly synergistic pathways by which toglobally affect protein kinases and thus membrane coupled functions. First, sincean increase of [Ca2+ ] ; can stimulate protein kinase activity in the absence ofsecond messengers (Hepler and Gilman 1992), a redistribution of intracellularCa2+ from mitochondrial (Tagawa et al. 1985) or endoplasmic stores (Rossier andPutney 1991) to the cytosol must be considered. However, as mentioned inChapter 1 [Ca 2+ ]; of anoxic turtle brain slices was found to remain at normoxiclevels - 0.2 pM (Bickler 1992). The possibility of a Ca 2+ -only mediated response104seems unlikely; however, hepatocyte [Ca 2+ ] ; in response to anoxia has not beenmeasured. Recent evidence demonstrating a 5 to 40% dose dependent (30 nM to10 /./M) increase in rat vagal motorneuron input resistance in response to the A 1 -receptor agonist cyclopentyladenosine strongly supports a role for adenosine inregulating membrane functions (Marks et al. 1993).Frameworks like that in Figure 12 are useful for guiding new experiments,and it is worth emphasizing that the turtle hepatocyte preparation described hereoffers an excellent opportunity to study the signal transduction pathways outlinedin this figure. Furthermore, since there are large scale changes in metabolism inresponse to anoxia (10 fold reduction in ATP turnover and 4 fold reduction in ionpumping), which make the effects of experimental perturbations more easilymeasured, the interplay between the many overlapping intracellular signals andsecond messengers ought to also be more easily recognized. In the concludingsection below, the kinds of studies and experiments suggested by the model arebriefly discussed.The dotted-line pathways outlined in Figure 12 can each be individuallyinvestigated and probed in a step-by-step fashion from extracellular receptor ligand-0 G protein interaction with adenylate cyclase -> changes in [Ca 2+ ] ; -> activationof protein kinases -> phosphorylation of glycolytic enzymes, membrane ionchannels, and ion pumps. The functional significance of these manipulations canthen be assessed by their effect on ATP turnover, measured from the rate oflactate accumulation during anoxia, or their effect on membrane permeability andion pump activity. This can be achieved through the use of specific inhibitors andactivators of protein kinases and G proteins, and receptor agonists and antagonists(see Chapter 1). For example, PKA can be inhibited by treating the hepatocyteswith IP20, then during anoxia the effect this has on PK and PFK phosphorylationcan be determined. One would predict that, inhibition of PK and PFK105phosphorylation would result in higher rates of lactate production than undercontrol non -treated anoxic cells. Also, the degree to which enzyme bindingeffects anoxic ATP turnover can be determined. The degree of enzyme binding tothe particulate fraction, measured by the homogenization of cells in low ionicstrength medium, and correlated with lactate production could ascertain theimportance of enzyme binding to the down-regulation of glycolysis. A morepowerful test of this effect would be to determine if mild oxidants, such as thoseused by Harrison et al. (1991) to regulate enzyme binding to erythrocyte band 3protein, could effect enzyme binding to the particulate fraction and alter lactateproduction.Phosphorylation of Na /K ATPase in response to anoxia by either PKA orPKC can be determined as described by Bertorello et al. (1991) through the use ofspecific protein kinase inhibitors and activators (outlined in Chapter 1). Briefly,PKA could be stimulated with dibutyryl cAMP or PKC with a diacylglycerolanalogue in normoxic hepatocytes and the rate of 32P-ATP then measured inpermeabilized cells. If Na + /K ATPase is phosphorylated in the presence ofprotein kinase stimulators, one would expect a decrease in Na + /IC ÷ ATPaseactivity. Conversely, PKA and PKC inhibitors such as IP20 and staurosporine couldbe applied to cells that are subsequently made anoxic. This should lead to anincreased anoxic Na + /K+ ATPase activity and lactate production rate. Ionchannel conductance and therefore cell membrane permeability can be bestmeasured using whole-cell patch-clamp techniques. Using this technique acutemembrane permeability changes in response to anoxia can be determined.Furthermore, if membrane permeability decreases, as would be expected from themodel, normoxic cells can be preincubated with PKA and PKC inhibitors to blockthe putative phosphorylation pathway during anoxia. If the framework is correct,this would then prevent a decrease in membrane permeability. Alternatively, PKA106and PKC stimulators (phorbol esters, diacylglycerol analogues) can be applied tonormoxic cells and changes in membrane permeability determined. During anoxia,protein kinase stimulators would not be expected to decrease membranepermeability further, since the model predicts that the protein kinases are alreadystimulated. The involvement of G proteins can also be assessed by use of the Gprotein inhibitor (pertusis toxin) or activator (cholera toxin) and the sametechniques as above applied to measure changes in glycolytic enzyme, ionchannel, and ion pump phosphorylation events. Inhibitory G protein activation,would be expected to prevent protein kinase activation during anoxia and blockthe expected decreases in glycolytic carbon flux, membrane ion permeability, andNa ÷/K+ ATPase activity. Activation of G protein, would be expected to stimulateprotein kinases and result in decreases in glycolytic carbon flux, membrane ionpermeability, and Na l" /K+ ATPase activity during normoxia. During anoxia, thiswould not be expected to have a great effect since as predicted by the modelprotein kinases would already be activated.Finally, these pathways can be probed at the extracellular level using a andadrenergic and adenosine receptor agonists (clonidine, isoproterenol, or 2-chloroadenosine, respectively; Exton 1985, Rudolphi et al. 1992) or antagonists(phentolamine, propanalol, or caffeine, respectively; Exton 1985, Rudolphi et al.1992) application. The responses to inhibitory or stimulatory ligands can bemeasured as above and also in combination with protein kinase and G proteinmodulators to dissect specific pathways.Therefore, in an anoxia tolerant hepatocyte preparation such as this theextracellular and intracellular signalling events, and underlying mechanisms leadingto a reduction of ATP production (metabolic arrest) and ATP utilizing membranecoupled functions (channel arrest) can be deduced.Literature CitedAndersson, B.S, T.Y. Aw, and D.P. Jones. 1987, Mitochondria! transmembranepotential and pH gradient during anoxia. Am. J. Physiol. 252:C349-C355,.Anundi, I., and H. De Groot. 1989, Hypoxic liver cell death: critical p02 anddependence of viability on glycolysis. Am. J. Physiol. 257:G58- G 64.Aw, T.Y. and D.P. Jones. 1989, Cyanide toxicity in hepatocytes under aerobicand anaerobic conditions. Am. J. Physiol. 257:C435-C441.Belkin, D.A. 1962, Anaerobiosis in diving turtles. Physiologist 5:105.Beres, L. and J.M. Sturtevant. 1971 Calorimetric studies of the activation ofchymotrypsinogen A. Biochem. 10:2120-2126.Bergmeyer, H.U. 1974, Methods in enzymatic analysis. Academic Press, NewYork.Bertorello, A.M. and A. Aperia. 1989, Na + -K + -ATPase is an effector protein forprotein kinase C in renal proximal tubule cells. Am. J. Physiol. 256:F370-F373.Bertorello, A.M., A. Aperia, S.I. Walaas, A.C. Nairn, and P. Greengard. 1991,Phosphorylation of the catalytic subunit of Na + ,K + -ATPase inhibits theactivity of the enzyme. Proc. Natl. Acad. Sci. USA. 88:11359-11362.Bickler, P.E. 1992, Cerebral anoxia tolerance in turtles: regulation of intracellularcalcium and pH. Am. J. Physiol. 263:R1298-R1302.Bickler, P.E. 1993, Cerebral hypoxia tolerance in turtles: role of glutamate receptorblockade. Lake Louise Symposium Proceedings.Billman, G.E. 1992, Cellular mechanisms for ventricular fibrillation. NIPS, 7:254-259.Bing, 0.H.L., W.W. Brooks, A.N. lnamdar, and J.V.Messer. 1972, Tolerance ofisolated heart muscle to hypoxia: turtle vs. rat. Am. J. Physiol.223(6):1481-1485.Bosca, L. and K.B. Storey. 1991, Inactivation of 6-phosphofructo-2-kinase duringanaerobiosis in the marine whelk Busycon canaliculatum. Am. J. Physiol.260:R1168-R1175.Bowen, J.W. 1992, Regulation of Na + -K + -ATPase expression in cultured renalcells by incubation in hypertonic medium. Am. J. Physiol. 262:C845-C853.Brezis, M., S. Rosen, K. Spokes, P. Silva, and F.H. Epstein. 1984, Transport-dependent anoxic cells injury in the isolated perfused rat kidney. Am. J.Pathol. 116:327-341.Brooks, S.P.J. and K.B. Storey. 1989, Regulation of glycolytic enzymes duringanoxia in the turtle Pseudemys scripta. Am. J. Physiol. 257:R278-R283.107Brooks, S.P.J. and K.B. Storey. 1993. Control of metabolic rate by multienzymecomplexes: is glycolysis in hypoxia and anoxia regulated by complexformation. pp. 281-293, in Surviving Hypoxia: Mechanisms of Control andAdaptation. eds P.W. Hochachka, P.L. Lutz, T. Sick, M. Rosenthal, and G.can den Thillart. by CRC press, Boca Raton FL.Brown, A.M. 1991, Ion channels as G protein effectors. NIPS, 6:158-161.Buck, L.T, S.C. Land, and P.W. Hochachka. 1993, Anoxia tolerant hepatocytes: Amodel system for the study of reversible metabolic suppression. Am. J.Physiol. in press.Buck, L.T. and P.W. Hochachka. 1993, An assessment of the importance ofsuccinate as an anaerobic endproduct in the diving turtle (Chrysemys pictabellii). J. Exp. Zool. submitted.Buck, L.T., P.W. Hochachka, A. SchOn, and E. Gnaiger. 1993, Microcalorimetricmeasurement of a large reversible anoxia induced metabolic suppression inisolated hepatocytes. Am. J. Physiol., in press.Busa, W.B. 1985, How to succeed at anaerobiosis without really dying. Mol.Physiol. 8(3):351-358.Busch, A.E., M.D. Varnum, R.A., North, and J.P. Adelman. 1992, An amino acidmutation in a potassium channel that prevents inhibition by protein kinase C.Science, 255:1705-1707.Castellini, M.A. and G.N. Somero. 1981, Buffering capacity of vertebrate muscle:Correlations with potentials for anaerobic function. J. Comp. Physiol.143:191-198.Chih, C-H., M. Rosenthal, and T.J. Sick. 1989, Ion leakage is reduced duringanoxia in turtle brain: a potential survival strategy. Am. J. Physiol.257:R1562-R1564.Clark, V.M. and A.T. Miller. 1973, Studies on anaerobic metabolism in the fresh-water turtle (Pseudemys scripta elegans), Comp. Biochem. Physiol. 44A:55-62.Cohen, P. 1992, Signal integration at the level of protein kinases, proteinphosphatases and their substrates. TIBS, 17:408-413.Cohen, B.J. and C. Lechene. 1990, Alanine stimulation of passive ion efflux inhepatocytes is independent of Na + -K ± pump activity. Am. J. Physiol. C24-C29.Curtain, N.A. and R.C. Woledge. 1978, Energy changes and muscular contraction.Physiol. Rev. 58(3):690-761.Daw, J.C., D.P. Wenger, and R.P. Berne. 1967, Relationship between cardiacglycogen and tolerance to anoxia in the western painted turtle, Chrysemyspicta bellii. Comp. Biochem. Physiol. 22:69-73.108109de Zwaan, A. 1983. Carbohydrate catabolism in bivalves. pp. 138-175 in TheMollusca ed. P.W. Hochachka, Academic Press, New York, NY.Doll, C.J., P.W. Hochachka, and P.B. Reiner. 1991, Effects of anoxia andmetabolic arrest on turtle and rat cortical neurons. Am. J. Physiol.260:R747-R755.Duncan, J.A. and K.B. Storey. 1992, Subcellular binding and the regulation ofglycolysis in anoxic turtle brain. Am. J. Physiol. 262(31):R517-R523.Edwards, R.A., P.L. Lutz, and D.G. Baden. 1989, Relationship between energyexpenditure and ion channel density in the turtle and rat brain. Am. J.Physiol. 257:R1354-R1358.Elliot, K.R.F, R. Ash, C.I. Pogson, S.A. Smith, and D.P. Crisp. 1976, Comparativeaspects of the preparation and biochemistry of liver cells from variousspecies. In Use of isolated liver cells and kidney tubules in metabolicresearch. Edited by J.P. Tager, Co., Amsterdam, The Netherlands. pp. 139-143.Else, P.L. and A.J. Hulbert. 1987, Evolution of mammalian endothermicmetabolism: "leaky" membranes as a source of heat. Am. J. Physiol.253:R1-R7.Erecinska, M. and I.A. Silver. 1989, ATP and brain Function. J. Cereb. Blood FlowMetab. 9:2-19.Farber, J.L., K.R. Chien, and S. Mittnacht. 1981, The pathogenesis of irreversibleinjury in ischemia. Am. J. Pathol. 102:271-281.Feng, Z., M. Rosenthal, and T. Sick. 1988, Suppression of evoked potentials withcontinued ion transport during anoxia in turtle brain. Am. J. Physiol.255(24):R478-R484.Fitz, J.G, T.E. Trouillot, and B.F. Scharschmidt. 1989, Effect of pH on membranepotential and K + conductance in cultured rat hepatocytes. Am. J. Physiol.257:G961-G968.Foster, G.D. and T.W. Moon. 1987, Metabolism in sea raven (Hemitripterusamericanus) hepatocytes: the effect of insulin and glucagon. Gen. Comp.Endocrin. 66:102-115.French, C.J., T.P. Mommsen, and P.W. Hochachka. 1981, Amino acid utilizationin isolated hepatocytes from rainbow trout. Eur. J, Biochem. 113:311-317.Gnaiger, E. 1980(a), Das kalorische Aquivalent des ATP-Umsatzes im aeroben andanoxischen Metabolismus. Thermochim. Acta 40:195-223.Gnaiger, E. 1980(b), Energetics of invertebrate anoxibiosis: direct calorimetry inaquatic oligochaetes. FEBS Lett. 112(2): 239-242.Gnaiger, E. 1983, Heat dissipation and energetic efficiency in animal anoxibiosis:economy contra power. J. Exp. Zool. 228:471-490.110Gnaiger, E. and R.B. Kemp. 1990, Anaerobic metabolism in aerobic mammaliancells. Information from the ratio of calorimetric heat flux and respirometricoxygen flux. Biochim. Biophys. Acta 1016:328-332.Goodyear, L.J., M.F. Hirshman, and E.S. Horton. 1991, Exercise-inducedtranslocation of skeletal glucose transporters. Am. J. Physiol. 261:E795-E799.Gorman Nordmark, M. J. Laynez, J. Suurkuusk, and I.Wadso. 1984, Design andtesting of a new microcalorimetric vessel for use with living cellular systemsand in titration experiments. J. Biochem. Biophys. Meth. 10:187-202.Gregory, G.A., F.A. Welsh, A.C.H. Yu, and P.H. Chan. 1990, Fructose -1,6-bisphosphate reduces ATP loss from hypoxic astrocytes. Brain Res.516:310-312.Hammerstedt, R.H., and H.A. Lardy. 1983, The effect of substrate cycling on theATP yield of sperm glycolysis. J. Biol. Chem. 258:8759-8765.Hansen, A.J. 1985, Effect of anoxia on ion distribution in the brain. Physiol. Rev.65:101-148.Hardewig, I., A.D.F. Addink, M.K. Grieshaber, H.0 Portner and G. van den Thillart.1991, Metabolic rates at different oxygen levels determined by direct andindirect calorimetry in the oxyconformer Sipuncu/us nudus. J. E. Biol.157:143-160.Harrison, M.L., P. Rathinavelu, P. Arese, R.L. Geahlen, and P. Low. 1991, Role ofband 3 tyrosine phosphorylase in the regulation of erythrocyte glycolysis. J.Biol. Chem. 266(7):4106-4111.Hepler, J.R. and A.G. Gilman. 1992, G Proteins. TIBS 17:383-387.Herbert, C.V. and D.C. Jackson. 1985(a), Temperature effects on the responsesto prolonged submergence in the turtle Chrysemys picta bellii. i. blood acid-base and ionic changes during and following anoxic submergence. Physiol.Zool. 58(6):655-669.Herbert, C.V. and D.C. Jackson. 1985(b), Temperature effects on the responsesto prolonged submergence in the turtle Chrysemys picta bellii. ii. Metabolicrate, blood acid-base and ionic changes, and cardiovascular function inaerated anoxic water. Physiol. Zool. 58(6):670-681.Herman, B., J.G. Gores, A-L. Nieminen, T. Kawanishi, A. Harman, and J.J.Lemasters. 1990, Calcium and pH in anoxic and toxic injury. Crit. Rev.Toxicol. 21(2):127-148.Hille, B. 1992. Ionic Channels of Excitable Membranes. 2ed Sinauer AssociatesINC. Publ. Sunderland, MA. pp. 170-201.Hindell, M.A., D.J. Slip, H.R. Burton, and M.M. Bryden. 1992, Physiologicalimplications of continuous, prolonged and deep dives of the southernelephant seal (Mirounga leonina). Can J. Zool. 70:370.111Hochachka, P.W., T.G. Owen, J.F. Allen and G.C. Whittow. 1975, Multiple endproducts of anaerobiosis in diving vertebrates. Comp. Biochem. Physiol.508:17-22.Hochachka, P.W. 1980, Living without oxygen, closed and open systems inhypoxia tolerance. Harvard University Press, Cambridge Ma.Hochachka, P.W. 1985, Assessing metabolic strategies for surviving 02 lack: roleof metabolic arrest coupled with channel arrest. Mol. Physiol. 8:331-350.Hochachka, P.W. 1986, Defense strategies against hypoxia and hypothermia.Science, 231:234-241.Hochachka, P.W. 1987, The mystery of the missing Pasteur effect in good animalanaerobes. Appl. Card. Pathophys. 1:33-43.Hochachka, P.W. and M. Guppy. 1987. Metabolic arrest and the control ofbiological time. Harvard University Press. Cambridge. MA. p. 180.Hochachka, P.W. and G.N. Somero. 1984. Biochemical Adaptation. Princeton, NJ:Princeton Univ. Press.Holwerda, D.A. and A. de Zwaan. 1979, Fumarate reductase of Mytilus edulis L.Mar. Biol.Lett. 1:33-40.Hue, L. 1982, Role of fructose-2,6-bisphosphate in the stimulation of glycolysis byanoxia in isolated hepatocytes. Biochem. J. 206:359-365.Hue, L. and M.H. Rider. 1987, Role of fructose 2,6-bisphosphate in the control ofglycolysis in mammalian tissues. Biochem. J. 245:313-324.Jackson, D.C. 1968, Metabolic depression and oxygen depletion in the divingturtle. J. Appl. Physiol. 24(4):503-509.Jackson, D.C. 1993. Anaerobic Metabolism in Reptiles. pp. 314-322. in TheVertebrate Gas Transport Cascade, Adaptations to Environment and Modeof Life. Ed. E.P.W. Bicudo. CRC Press, Boca Raton, FL.Jackson, D.C., C.V. Herbert, and G.R. Ultsch. 1984, The comparative physiologyof diving in North American freshwater turtles. ii. plasma ion balance duringprolonged anoxia. Physiol. Zool. 57(6):632-640.Jackson, D.C. and G.R. Ultsch. 1982, Long-term submergence at 3°C of the turtleChrysemys picta bellii, in normoxic and severely hypoxic water: Extracellularionic responses to extreme lactic acidosis. J. Exp. Biol. 96:29-43.Jackson, D.C. and N. Heisler. 1982, Plasma ion balance of submerged anoxicturtles at 3°C: The role of calcium lactate formation. Resp. Physiol. 49:159-174.Jackson, D.C. and N. Heisler. 1983, Intracellular and extracellular acid-base andelectrolyte status of submerged anoxic turtles at 3°C. Resp. Physiol.53:187-201.112Keiver, K.P., J. Wienberg, and P.W. Hochachka. 1992, The effect of anoxicsubmergence and recovery on circulating levels of catecholamines andcorticosterone in the turtle, Chrysemys picta. Gen. Comp. Endocrinol.72:63-71.Kelly D.A. and K.B. Storey. 1988, Organ-specific control of glycolysis in anoxicturtles. Am. J. Physiol. 255:R774-R779.Keppens, S., J.R. Vandenheede and H. De Wulf. 1977, On the role of calcium assecond messenger in liver for the hormonally induced activation of glycogenphosphorylase. Biochim. Biophys. Acta 496;448-457.Kinter, D., J.H. Fitzpatrich, Jr., J.A. Louie, and D.D. Gilboe. 1984, Cerebral 02and energy metabolism during and after 30 minutes of moderatehypoxia. Am. J. Physiol. 247:E475-E482.Land, S.C., L.T. Buck and P.W. Hochachka. 1993, Response of protein synthesisto anoxia and recovery in anoxia tolerant hepatocytes. Am. J. Physiol (inpress).Lemasters, J.J., J. DiGuiseppi, A. Nieminen, and B.Herman. 1987, Blebbing, freeCa2+ and mitochondrial membrane potential preceding cell death inhepatocytes. Nature 325:78-81.Lemasters, J.J., J.M. Bond, R.T. Curtain, A.-L. Niemininen, J.C. Caldwell-Kenkel,D.C. Harrison, S.H. Kaplan, W.E. Cascio, R.G. Thurman, G.J. Gores, and B.Herman. 1993, Reperfusion injury to heart and liver cells: Protection byacidosis during ischemia and a "pH Paradox" after reperfusion. in SurvivingHypoxia: Mechanisms of Control and Adaptation. eds P.W. Hochachka, P.L.Lutz, T. Sick, M. Rosenthal, and G. can den Thillart. by CRC press, BocaRaton FL.Londos, C., D.M.F. Cooper, and J. Wolff. 1980, Subclasses of external receptors.Proc. Natl. Acad. Sci. USA. 77(5):2551-2554.Lutz, P.L. 1992, Mechanisms for anoxic survival in the vertebrate brain. Annu.Rev. Physiol. 54:601-618.Lutz, P.L., P. McMahon, M. Rosenthal, and T. Sick. 1984, Relationships betweenaerobic and anaerobic energy production in turtle brain in situ. Am J.Physiol. 247(16)R740-R744.Lutz, P.L., M. Rosenthal, and T.J. Sick. 1985, Living without oxygen: turtle brainas a model of anaerobic metabolism. Mol. Physiol. 8:411-425.Mommsen, T.P. and K.B. Storey. 1992, Hormonal effects on glycogen metabolismin isolated hepatocytes of a freeze-tolerant frog. Gen. Comp. Endocrin.87:44-53.Moon, T.W., P.J. Walsh, and T.P. Mommsen. 1985, Fish hepatocytes: a modelmetabolic system. Can. J. Aquat. Sci. 42:1772-1782.Morris, G.M. and J. Baldwin. 1984, pH buffering capacity of invertebrate muscle:correlations with anaerobic muscle work. Mol. Phyiol. 5:61-70.Nayler, W.G. 1983, Calcium and cell death. Eur. Heart J. 4:33-41.Nilsson, G.E. 1991. The adenosine receptor blocker aminophyilline increasesanoxic excretion in crucian carp. Am. J. Physiol. 261(30):R1057-R1060.Nilsson G.E. and P.L. Lutz. 1992, Adenosine release in the anoxic turtle brain: Apossible mechanism for anoxic survival. J. Exp. Biol. 162:345-351.Nobes, C.D. and M.D. Brand. 1989, A quantitative assessment of the use of 360"distribution to measure plasma membrane potential in isolated hepatocytes.Biochim. Biophys. Acta, 987:115-123.Nobes, C.D., P.L. Lakin-Thomas, and M.D. Brand. 1989, The contribution of ATPturnover by the Na + /K ± ATPase to the rate of respiration of hepatocytes.Effects of thyroid status and fatty acids. Biochim. Biophys. Acta 976:241-245.Numann, R., W.A. Caterall, and T. Scheuer. 1991, Functional modulation of brainsodium channels by protein kinase C phosphorylation. Science, 254:115-118.Penney, D.G. 1974, Effects of prolonged diving anoxia on the turtle, Pseudemysscripta elegans. Comp. Biochem. Physiol. 47A:933-941.Penney, D.G. and E.H. Kornecki. 1973, Activities, intracellular localization andkinetic properties of phosphoenolpyruvate carboxykinase, pyruvate kinase,and malate dehydrogenase in turtle (Pseudemys scripta elegans) liver, heart,and skeletal muscle. Comp. Biochem. Physiol. 46B:405-415.Perez-Pinz•in, M.A., C.Y. Chan, M. Rosenthal, and T.J. Sick. 1992(a), Membraneand synaptic activity during anoxia in the isolated turtle cerebellum. Am. J.Physiol. 263:R1057-R1063.PerOz-PinzOn, M.A., M. Rosenthal, T.J. Sick, P.L. Lutz, J. Pablo, and D. Mash.1992(b), Downregulation of sodium channels during anoxia: a putativesurvival strategy of turtle brain. Am. J. Physiol. 262(31):R712-R715.P6rOz-Pinzon, P.A., P. Rosenthal, P.L. Lutz, and T.J. Sick. 1992(c), Anoxicsurvival of the isolated cerebellum of the turtle Pseudemys scripta elegans.J. Comp. Physiol. B. 162:68-73.Plaxton, W.C. and K.B. Storey. 1984, In vivo phosphorylation of red musclepyruvate kinase from the channeled whelk, Busycotypus canaliculatum, inresponse to anoxic stress. Eur. J. Biochem. 143:267-272.Plaxton, W.C. and K.B. Storey. 1986, Glycolytic enzyme binding and metaboliccontrol in anaerobiosis. J. Comp. Physiol. B, 156:635-640.Robin, E.D., N. Lewiston, A. Newman, L.M. Simon, and J. Theodore. 1979,Bioenergetic pattern of turtle brain and resistance to profound loss ofmitochondrial ATP generation. Proc. Natl. Acad. Sci. USA. 76(8):3922-3926.113114Rossier, B.C., K. Geering, and J.P. Kraehenbuhl. Regulation of the sodium pump:how and why? TIBS 12:483-487, 1987.Rottenberg, H. 1979, The measurement of membrane potential and •pH in cells,organelles, and vesicles. Meth. Enzymol. 15:547-586.Rudolphi, K.A., P. Schubert, F.E. Parkinson, and B.B. Fredholm. 1992, Adenosineand Brain Ischemia. Cerebrovasc. Brain Metab. Rev. 4:346-369.Satoh, T., H.T. Cohen, and A.I. Katz. 1992, Intracellular signalling in theregulation of renal Na-K-ATPase. I. Role of cyclic AMP and phospholipaseA2. J. Clin. Invest. 89:1496-1500.Schmidt-Nielsen, K., 1984, Animal Physiology, 2nd ed. Cambridge, CambridgeUniversity Press, p. 207.Sch6n, A. and I. Wads& 1986, Thermochemical characterization of T-lymphomacells under non-growing conditions. Cytobios 48:195-205.Schubert, B., A.M.J. Van Dongen, G.E. Kirsh, and A. Brown. 1989, 13-Adrenergicinhibition of cardiac sodium channels by dual G-protein pathways. Science,245:516-519.Schulte, P.S., C.D. Moyes, and P.W. Hochachka. 1992, Integrating metabolicpathways in post-exercise recovery of white muscle. J. Exp. Biol. 166:181-195.Scott, I.D. and D.G. Nicholls. 1980, Energy transduction in intact synaptosomes.Biochem. J. 186:21-33.Shick, J.M., A. de Zwaan, and A.M.Th. de Bont. 1982, Anoxic metabolic rate inthe mussel Mytilus edulis L. estimated by simultaneous direct calorimetryand biochemical analysis. Physiol. Zool. 56(1):56-63.Shick, J.M., J. Widdows, and E. Gnaiger. 1988, Calorimetric studies of behaviour,metabolism, and energetics of sessile intertidal animals. Amer. Zool.28:161-181.Sick, T.J., M. Perez-Pinzon, P.L. Lutz, and M. Rosenthal. 1993. Maintainingcoupled metabolism and membrane function in anoxic brain: A comparisonbetween the turtle and rat. pp. 351-363, in Surviving Hypoxia: Mechanismsof Control and Adaptation. eds P.W. Hochachka, P.L. Lutz, T. Sick, M.Rosenthal, and G. can den Thillart. by CRC press, Boca Raton FL.Siebenaller, J.F. 1979, Regulation of pyruvate kinase in Myti/us edulis byphosphorylation-dephosphorylation. Mar. Biol. Lett., 1:105-110.Sigel, E. and R. Baur. 1988, Activation of protein kinase C differentially modulatesneuronal Na +, Ca +, and gamma-aminobutyrate type A channels. Proc.Natl. Acad. Sci. USA. 85:6192-6196.Shimizu, H. 1983, Adenosine receptors associated with the adenylate cyclasesystem. pp. 31-40, in Physiology and Pharmacology of Adenosine115Derivatives, eds. J.W. Daly, Y. Kuroda,H. J.W. Phillips, Shimizu, and M. Ui.Raven Press, New York.Skou, J.C. 1992, The Na-K pump. NIPS 7:95-100.Srere, P.A. 1987, Complexes of sequential metabolic enzymes. Annu. Rev.Biochem. 56:89-124.Storey, K.B. and J.M. Storey. 1990, Metabolic rate depression and biochemicaladaptation in anaerobiosis, hibernation and estivation. Quart. Rev. Biol.65(2):145-174.Storey, K.B. 1987, Tissue specific controls on carbohydrate catabolism duringanoxia in goldfish. Physiol. Zool. 60:601-607.Suarez, R.K., and T.P. Mommsen. 1985, Gluconeogenesis in teleost fishes. Can.J. Zool. 65:1869-1882.Suarez, R.K., P.D. Mallet, C. Daxboeck, and P.W. Hochackha. 1985, Enzymes ofenergy metabolism and gluconeogenesis in the pacific blue marlin, Makairanigricans. Can. J. Zool. 64:694-697.Sutherland, E.W. 1971, Studies on the mechanisms of hormone action. Science,177:401-408.Taetmeyer, H. 1978, Metabolic responses to cardiac hypoxia. Increasedproduction of succinate by rabbit papillary muscles. Circ. Res. 43(5):808-815.Tewari, Y.B., D.K. Steckler, and R.N. Goldberg. 1988, Thermodynamics ofhydrolysis of sugar phosphates. J. Biol. Chem. 263(8):3670-3675.Tolbert, N.E. and E. Essner. 1981, Microbodies: Peroxisomes and Glyoxysomes.J. Cell Biol. 91(3):271s-283s.Tria, E., P. Luly, V. Tomasi, A. Trevisani, and 0. Barnabei. 1974, Modulation bycyclic AMP in vitro of liver plasma membrane (Na ± -K + )-ATPase and proteinkinases. Biochim. Biophys. Acta, 343:297-306.Ultsch, G.R., and D.C. Jackson. 1982, Long-term submergence at 3°C of theturtle Chrysemys picta bell, in normoxic and severely hypoxic water. i.Survival, gas exchange and acid-base status. J. Exp. Biol. 96:11-28.Ultsch, G.R., C.V. Herbert, and D.C. Jackson. 1984, The comparative physiologyof diving in North American freshwater turtles. i. Submergence tolerance,gas exchange, and acid-base balance. Physiol. Zool. 57(61:620-631.van den Thillart, G. and A. van Waarde. 1985, Teleosts in hypoxia: aspects ofanaerobic metabolism. Mol. Physiol. 8(3):393-409.van Waarde, A. 1991, Alcoholic fermentation in multicellular organisms. Physiol.Zool. 64(4):895-920.116van Waversveld, J., A.D.F. Addink, and G. van den Thillart. 1989, Simultaneousdirect and indirect calorimetry on normoxic and anoxic goldfish. J. Exp.Biol. 142:325-335.Yatani, A., J. Codina, Y. Imoto, J.P. Reeves, L. Birnbaumer, and A. Brown. 1987,A G-protein directly regulates mammalian cardiac calcium channels.Science, 238:1288-1292.Walsh, P.J. 1989, Regulation of intracellular pH by toadfish (Opsanus beta)hepatocytes. J. Exp. Biol. 147:407-419.Wasser, J.S., K.C. Inman, E.A. Arendt, R.G. Lawler and D.0 Jackson. 1990, 31 P-NMR measurements of pHi and high-energy phosphates in isolated turtlehearts during anoxia and acidosis. Am.J. Physiol. 259:R521-R530.Wasser, J.S., E.A. Meinertz, S.Y. Chang, R.G. Lawler and D.0 Jackson. 1992,Metabolic and cardiodynamic responses of isolated turtle hearts to ischemiaand reperfusion. Am. J. Physiol. 262:R437-R443.Wasser, J.S., S.J. Warburton, and D.C. Jackson. 1991, Extracellular andintracellular acid-base effects of submergence anoxia and nitrogen breathingin turtles. Resp. Physiol. 83:239-252.


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