Kidney International, Vol. 51 (1997), PP. 386—392

Is tubuloglomerular feedback a tool to prevent nephron oxygendeficiency?

HANS-JOACHIM SCHUREK and OLIVER JOHNS

Medizinische Hochschule Hannover, Zentrum Innere Medizin, Abteilung Nephrologie, Hannover, and Nephrologisches Zentrum Emsland am St.Bonifarius Hospital, Akademisches Lehrkrankenhaus der MHH, Lingen, Germany

Is tubuloglowerular feedback also a tool to prevent nephron oxygendeficiency? The purpose of the study was to analyze whether rhythmicoscillations of proximal tubular pressure and distal fluid conductivity at therenal surface are induced by oxygen deficiency of thick ascending limb(TAL) segments. Oxygen pressure was measured in halothane anesthe-tized Munich-Wistar rats by a multi-wire micro-gold electrode at thekidney surface. Signals from wires placed upon glomeruli and tubuliexhibited P°2 oscillations with exactly the same frequency (in mean 30mHz) as have been described for proximal tubular pressures or distal fluidconductivities. This supports our suggestion that a limited oxygen supplyto the nephron forces TAL segments to oscillate between aerobic andanaerobic energy production. A switch to glyeolysis reduces TAL's trans-port efficiency dramatically. At the macula densa, the terminal end of theTAL segment, the thereby elevated sodium concentration operates as aswitch by means of the TGF to adapt the filtered load to the oxygen supplyof the individual nephron. In this way proximal tubules may also beprotected from oxygen deficieny, which is essential due to their lowglycolytic capacity. An enhanced halothane concentration of 2% or the useof barbiturates, such as mactin®, blocks oscillations completely as furo-semide blocks oscillations as well as the feedback response. Reduction ofthe hematocrit by exchange transfusion mainly reduces supratubular P°2values, and to a lesser extent also reduces supraglomerular pressures. Thisdemonstrates that oxygen shunt diffusion in the kidney cortex and medullais a prerequisite for both the function of a sensor to measure P°2 andoxygen capacity to regulate erythropoietin Secretion and to enable aneffective adjustment of blood flow to the metabolic and functionaldemands of the kidney.

Renal blood flow is high, but not luxuriously high as it has beenthought for decades; it results from the metabolic demand of bothrenal medulla and cortex. Within both areas an effective shuntdiffusion is present for blood gases [1—3], trapping CO2 to reachelevated levels within the cortex [4, 5] and the papilla, whichfacilitates the shunt diffusion of oxygen by the Bohr effect. Shuntdiffusion reduces P°2 to the brink of oxygen deficiency in parallel[6—8]. Within the outer medulla, oxygen extraction is surprisinglyhigh 16] and thereby the oxygen supply of the thick ascending limb(TAL) segments, especially those of the outer cortical nephrons,is limited [7, 9]. More interesting is the fact that TAL segmentshave a high mitochondrial volume density [10] and also a highcapacity for glycolysis [11, 121. The tubuloglomcrular feedback(TGF) mechanism may be a perfect tool to adapt the tubularenergy demand to its actual oxygen supply 113]. If the filteredtubular fluid and sodium-chloride load exceeds the reabsorption

© 1997 by the International Society of Nephrology

capacity of this individual nephron, the TAL segment transformsthe enhanced fluid volume into an increased salt concentration[14], which at the macula densa cells can be used to shut down thefiltered load via TGF. This adaptation operates in an oscillatingfashion, experimentally in the rat with a frequency of 30 mHz(2/mm) and can be measured by the oscillation of proximaltubular pressure [14, 15], an oscillation of distal sodium concen-tration (phase shifted), and also by the oscillation of P°2 asmeasured at the renal surface [16] by a multi-wire micro-goldelectrode 117]. The high respiratory capacity of the TAL segmentscombined with an oxygen supply at the brink of deficiency and ahigh glycolytic enzyme activity are thought to function as a switchbetween the aerobic energy supply and anaerobic glycolysis.During glycolysis ATP production declines drastically from 24 to3 M/M glucose and sodium concentration at the terminal end of theTAL segment, the site of the macula densa is increased. At thispoint the work load of the whole nephron can be adjusted to theavailable energetic stores (oxygen). Thereby the TOF mechanismcan induce oscillations and also protect proximal tubules fromoxygen deficiency, which is essential due to their low anaerobiccapacity.

Methods

Twenty-seven male Munich-Wistar-Frömter rats (MWF) wereused in a range of 250 to 380 g body wt. This MWF strain wasselected for its high numbers of superficial glomeruli [3]. Theanimals had free access to tap water and 12 hours before theexperiment to a pelleted standard diet (Altromin 1310). Theywere anesthetized with halothane and artificially ventilated via atracheal cannula, placed on a temperature-controlled operationplatform to keep body temperature at 3TC as measured by arectal probe. The right femoral artery and vein were cannulatedby short PE-50 tubes connected to Tygon tubing, establishing anarterio-venous shunt. The shunt was connected to a thcrmostattedglass cuvette housing a Clark-02 electrode (Eschweiler, Kid,Germany). The shunt tubes were connected to a Perspex blockcarrying siliconized steel inlet ports to provide multiple access forserum infusion, heparin infusion, and an access to flush the °2sensor cuvette. During the experiment, rat serum diluted withRinger lactate and isotonic glucose solution (25 vol% each) wasinfused at a rate of 2 to 3 ml/hr, as detailed elsewhere [3, 13]. Theexposed left kidney was immobilized within a thermostattedchamber and surrounded by moistened cotton-wool (Fig. 1). The

386

bubble treecalibration

Schurek and Johns: TGF prevents nephron oxygen deficiency 387

Fig. 1. Schematic diagram of the experimentalsetup: (1) thermostat-controlled rat kidney tray;(2) A V-shunt, a/v. femoralis; (3) mini-heatexchanger; (4) perspex block carlying inlet ports;(5) thermostat-controlled glass cuvette for a P°2electrode; (6) tube for (7) blood pressuremeasurement and (8) infusion line for heparin;(9) tube for substitution fluid and blood probes;(10) double syringe pump for exchangetransfusion; (11) small animal respirator; (12)warming up and moistening of the inspired gas;(13) halothane vaporizer; (14) gas mi.sing pump(Wösthoffi Bochum, Germany).

25

20

15

c'j010

5Fig. 2. Original calibration curve taken from an8-wire micro-gold electrode. The small insertgives a scheme of the thermostat-controlled (1

0 & 2, water inlet and outlet) glass calibration0 10 20 30 40 50 chamber designed to keep gas bubbles away

from the electrode during calibration (3 & 4,Time, minutes

gas inlet and outlet; 5, sintered glass).

multi-wire micro-electrode with a diameter of 5 mm has been adjusting a counter-weight. For calibration a specially designeddeveloped by Metzger, Hartmann and Wadouh (detailed in [171) glass cuvette was used (Fig. 2).and produced in our workshop (Medizinische Hochschule Han-nover) using an Ag/AgCI-cathode as reference electrode in the Isovolemic exchange transfusioncenter (200 jxm) surrounded by 8 gold wires of 15 jxm diameter A conventional syringe pump has been converted to thiseach embedded in Hysol (Dexter Corp., NY, USA). The dee- purpose (Fig. 1). The amount of blood withdrawn by the av-shunttrode was mounted on a specially designed micromanipulator to was replaced simultaneously by serum drawn from a littermate orthe arm of a record player to reduce its weight to a minimum by by a packed red cell suspension with a hematocrit between 75to

5 4

11

14

t

It

0— recording1.—i-- -h r

388 Schurek and Johns: TGF prevents nephron oxygen deficiency

Fig. 3. Recording of p02 at the surjiice of anexposed rat kidney with an 8-wire micro-goldelectrode, the geometiy of which is shown in theinsert (top left in topographic relation to surfaceglomeruli and tubuli). Inspiratory P°2 waschanged from room air to 95% °2/% CO., andvice versa. Halothane concentration was heldconstant at 0.8%, systemic P°2 was initially 80mm Hg, then 600 mm Hg and back to 90 mmHg. Simultaneous recording on paper wasrestricted to four channels, but could beswitched from K1-K4 to K5-K8. Starting pointsof switching P°2 have been marked, phaseshifts due to the recorder have beenconsidered: glomerular P°2 increases withouttime lag, and tubular P°2 increases after a timelag of about eight seconds ( time). Theincrease of P°2 above the glomerular field washigher and the slope steeper compared to thetubular field.

85%. The substitute passed a mini-heat exchanger. Nine animalswere used for this type of experiment.

ResultsAt a systemic P02 in a mean of 100 mm Hg and a hematocrit of

40% we could measure the following mean oxygen pressure valuesat the renal surface: supratubular values: 37.5 6.2 mm Hg (xSD, N = 49) and supraglomerular values: 55.2 3.0 mm Hg (N =11). For comparison under the anesthesia with mactin® thesupratubular value was 32.8 9.4 mm Hg (N = 19), andsupraglomerular was 54.9 2.5 mm Hg (N = 8).

The mean frequency of oscillations was 30 2 mHz (range 1.5to 2.2/mm). At a systemic P°2 of 100 mm Hg (maximal range 70to 130 mm Hg) the oscillation amplitude of supraglomerularvalues was 3.9 2.4 mm l-Ig (N = 52, range I to 10), andsupratubular values was 3.3 2.0 mm Hg (N = 45, range ito 8).At a systemic P°2 of 600 mm Hg (550 to 650 mm Hg) theamplitudes were for glomerular, 15 7.6 mm Hg (N = 26, range5 to 30) and tubular, 8.6 4.7 mm Hg (N 28, range 2 to 20).Oscillations as measured within a geometric field of ca. 400 xm2were all synchronized, that is, the supratubular values shifted inphase to supraglomerular values. An increase in systemic P°2 wasfollowed by a prompt supraglomerular increase, whereas tubularvalues showed a time lag of ca. 5 to 8 seconds. The glomerularincrease was more pronounced. Figure 3 gives a recording of fourwires of the 8-wire surface electrode, which includes a switch from

a systemic arterial P°2 of 80 mm Hg to 600 mm Hg and back to90 mm Hg. Figure 4 gives three examples of oscillations underdifferent experimental conditions.

The frequency of oscillations was dependent on the systemicP°2' ranging from 0% at values < 70 mm Hg, 10% of values at 70to 130 mm Hg, and 100% at the respiration of pure oxygen (P02= 600 mm Hg). A high concentration of halothane at> 2% blocksthe oscillations as effectively as it was seen under mactin® underall conditions (not shown).

Table 1 gives an overview of tubular P°2 when the hematocritwas varied by exchange transfusion replacing either serum orpacked red cells. At a mean Hct of 38% there is a clear cutrelationship between systemic p02 and tubular P°2 In compari-son, the increase of tubular P°2 in relation to systemic P°2 isdepressed at a low hematocrit.

Discussion

The ability to produce a concentrated urine is imposed by anuniquely low ambient oxygen pressure in the renal medulla due toshunt diffusion within the vascular bundles. As the TAL-segmentis also able to glycolyse anaerobically, a phase of oxygen deficiencycan he bridge-spanned. It allows an exceptionally high oxygenextraction of 80% in this area [6J. Phases of hypoxia moreover arethought to induce gene expression for glycolytic enzymes [18]. Ifoxygen capacity is reduced systematically, which can be effected inthe isolated kidney model by using cell free perfusate, a typical

Schurek and Johns: TGF prevents nephron oxygen deficiency 389

AP°2' mmHg b1 room airIIl95% 02/5% 002-----I--100 — — — — —

vtEF I

Fig. 4. Oscillations of p02 at the renal su,face at various conditions. Rats have been anesthetized by halothane 0.8% from beginning to end. At the topright of the figures the topographic positions of channels K1-4 and 1(5-8 are characterized. (A) Systemic arterial P°2 during inspiration of room air: 100mm Hg with a rapid change to inspiration of 95% 02/5% CO2. (B) Oscillations of P°2 after a short segmental ischemia, recording at the previouslyischemic field. Change of oxygen pressure. (C) Oscillations of P°2 during an isovolemic hemodilution. Tubular P°2 (K3, K4) shows a decrease in themean, whereas the glomerular PO2 (Ki, K2) is constant or slightly increases.

Table 1. Mean tubular P°2 at renal surface during exchange transfusion at various systemic P°2

Data are mean values SD.

pattern of lesions occurs in TAL segments and P3 segments [7], of the distal nephron [191 to protect the organism from losing saltand the reason for the development of these lesions may lie in a and water, but also to protect primarily proximal tubuli and TALnonfunctioning tubuloglomerular feedback (TGF) mechanism, as segments from oxygen deficiency. The demonstration of a highlydemonstrated in this preparation. Vice versa, the function of an effective cortical shunt diffusion for blood gases, which can beintact TGF mechanism in vivo can be seen as restricting the localized mainly to the interlobular arteries [31, which arc sur-amount of NaCI escaping into the low-capacity transport system rounded by thin-walled interlobular veins [20], and the frequency

B

H

t95%02/5%C02

C

4)

room air

— —C3

I—

c—

F-

yst. art. P°2130mm H

I

80- I4- I—

1

El

80

I-—-— —

70 I 1--

) —

—a,y k

60 — I1'iI40

50

I40

-— h... -. —

-I-I 20

1 mm. -

C'

— recording—1 min.—

— recorthng fr

N

HematocritSystemic

P°2 16 20—26 26—30 31—35

(%) range

36—40 41—45 46—50 52

50 17±310

60 19±7 23±3N 10 21

70 17±5 23±6 24±5 26±6N 9 16 22 12

80 25±7 22±4 28±5 30±6N 10 6 19 49

90 29±5 36±4 34±6N 7 13 39

100 30±7 31±9 36±4 38±6N 7 7 15 49

110 41±6N 28

120 25±4 35±9 32±5 36±7 44±5N 12 16 8 23 20

32 630

30 22

35 43

46 74

39 89

46 718

Fig. 5. Graphic course of (A) A TP and (B)glucose stores in hypoxia in proximal convoluted

120 tubules (0) and thick ascending limb (LI)segments. Graph drawn from data published byBastin et at [301 (used with permission).

distribution of P°2 values with a high fraction of low values withinthe renal cortex [3, 211 may be the main reason for a criticaloxygen supply to also be within the renal cortex. It seems to beessential to have an effective defense mechanism to preventoxygen deficiency in proximal tubular cells by a restriction oradaptation of the proximal work load. Another byproduct of thiseffective regulation seems to be the production of erythropoietin,which has been localized to the peritubular fibroblast-like cells inthe renal cortical interstitium [22], where a sensor mechanism hasbeen claimed to be localized for the sensing of P°2 and oxygencapacity. This sensor may also function as a measure of the overallmetabolic activity of the organism. A reduction of the oxygentransport capacity, as effected experimentally (Table 1) by thereduction of hematocrit, results in a proportional decrease oftubular P°2, which can also be used to regulate the renal corticalproduction of erythropoietin.

The easy way to anesthetize small laboratory animals by barbi-turates has masked the phenomenon of functional oscillations oftubular transport activity for a long time. It was first shown byLeyssac and Baumbach [151 when they used halothane instead ofthiopental according to the description of Kaczmarczyk andRheinhardt [23]. Since that time, several reports have beenpublished, and it is now established that the periodic oscillationarises in the operation of TGF [14, 15, 24]. These results pavedthe way for our own work [25, 261. It is a challenge to explain whyTAL segments exhibit a high respiratory capacity (mitochondrialdensity) and a high glycolytic capacity as well. It seems likely thatthis biochemical feature is a prerequisite for the switch function toadapt oxygen supply to oxygen demand. The strategic position ofthe TAL segments in front of the macula densa cells, which arcoperator cells for the TGF, make them to the ideal segmentcontrolling the nephron work load. Oscillatory phenomena arewell known in biochemistry [27, 281 and a synchronous metabolicrate of a given cell population is often initiated by suddenmetabolic pertubation, such as aerobic-anaerobic transition oraddition of oxygen or substrate, forcing the individual metabolicsystem of the population to identical states [28]. It has been shownby Holstein-Rathlou, that a suhpopulation of nephrons can syn-chronize their oscillations when they arise from the same inter-

lobular artery, and he coined it "cross-talk" [14]. It is now evidentthat oscillations are transmitted via the interlobular artery fromone afferent arteriole to another neighboring one [29], otherwiseit would also be difficult to explain why supraglomerular P°2values oscillate with an even higher amplitude than supratubularones.

The time course of ATP and glucose stores in hypoxia inproximal convoluted tubules (PCT) and TAL segments as shownin Figure 5 has been drawn from data published by Bastin et al[301. It demonstrates the rapid loss of ATP in PCT, but a longlasting reserve in TAL segments due to a rapid loss of glucose,which is thought to have been glycolysed to held ATP stores highfor some 60 seconds. This time course corresponds well to thefrequency of oscillations shown in the nephron function.

Figure 6 gives a synoptic presentation of all parameters, whichhave been shown to oscillate and as they have been publishedduring the last decade and used in the present discussion [13—15,24, 27, 28, 30]. In the biochemical literature, especially theconcentrations of glycolytie substrates that are coupled to theadenosine phosphate system [28] that may oscillate, the ATPsystem obviously is responsible for the propagation of the oscil-lation along the metabolic pathway or chain.

This represents a new dimension for the significance of theTGF and a prerequisite for the regulation of erythropoietinproduction within the renal cortex, which is far from beeingluxuriously supplied with oxygen.

Acknowledgments

This work has been supported by the Deutsche Forschungsgemeinsehaftgrant Schu 343/6-1. The group in Hannover was also supported by theState of Lower Saxony, Department of Science and Arts. Parts of this workhave been presented in abstract form at the American Society of Ncphrol-ogy annual meeting [161. Oliver Johns published his thesis in 1994 [13].The authors are grateful to Jorg Dieter Biela for technical assistance andgraphics work. We acknowledge the support, encouragement and discus-sion of A. Kurtz, Ch. Bauer, D.W. Lubbers, W. Kriz and B. Kaissling. H.Metzgcr in Hannover made it possible for us to use the P°2 micro-electrodes as developed in the Hannover workshop as well as the amplifierfor signal recording. The micromanipulator was also manufactured in ourworkshop after a special design, as well as a new type of calibration

390 Schurek and Johns: TGF prevents nephron oxygen deficiency

A15

10

0

0 30 60 90 120 0 30 60 90

Time. seconds Time, seconds

/ ----I - - (3a) 09

I

H

_____________________ (6Hww)

N ' N N N N

N 'S N \ 'I

'S N

N N N \ 5' N

N N N N \ N

N N rvi N N1 N_

5'

0

I' ————'

/ /-

'S 'S / /

Fig. 6. Synoptic presentation of a nephron exhibiting oscillations of proximal tubular hydrostatic pressure P.1 oscillation of distal fluid conductivity (Na )orchloride concentration and synchronous oscillations of surface P°2 above the tubular and the glomerular field. Therefore, sodium load at the beginningof the TAL-segment oscillates in the same frequency presumably as an oscillation of tubular flow, and the TAL-segment transforms this flow signal intoa change in salt concentration, the more so when metabolism switches from aerobic to anaerobic ATP production. The oscillating sodium concentrationat the niacula densa is able to use the TGF to adapt tubular transport to oxygen supply of the nephron and thus may be the driving forcc for the originof this type of oscillation.

392 Schurek and Johns: TGF prevents nephron oxygen deficiency

cuvette, which avoids the accumulation of gas bubbles at the electrodemembrane by a Ussing chamber-like gassing.

Reprint requests to Prof Dr. Hi. Schurek, Nephrologisches ZentrumEmsiand am St. Bonifalius-Hospital, Akademisches Lehrkrankenhaus derMedizinischen Hochschule Hannover, Poslfàch 1832, D-49788 Lingen/Ems,Germany.

References

I. LEVY MD, IMPERIAL ES: Oxygen shunting in renal cortical andmedullary capillaries. Am J Physiol 200:159—162, 1961

2. AUKLAND K, BOWER BF, BERLINER RW: Measurement of local bloodflow with hydrogen gas. Circ Res 14:164—187, 1964

3. SCHUREK HJ, JOST U, BAUMGARTL H, BERTRAM H, HECKMANN U:Evidence for a preglomerular oxygen diffusion shunt in rat renalcortex. Am J Physiol 259 (Renal Fluid Electrol Physiol 28):F910—F9 15,1990

4. SOHTELL M: CO2 along the proximal tubules in rat kidney. Acta ScandPhysiol 105:146—155, 1979

5. DUBOSE TD, CAFLISH CR, BIDANI A: Role of metabolic CO,-production in the generation of elevated renal cortical pCO2. Am JPhysiol 246:F592—F599, 1984

6. BREZIS M, ROSEN S, SILVA P, EPSTEIN FF1: Renal ischemia: A newperspective. Kidney mt 26:375—383, 1984

7. SCHtJREK Hi, KRIZ W: Morphologic and functional evidence foroxygen deficiency in the isolated perfused rat kidney. Lab Invest53:145—155, 1985

8. BAI.ABAN RS, SILVA AL: Spectrophotometric monitoring of O2delivery to the exposed rat kidney. Am J Physiol 241 :F257—262, 1981

9. EPSTEIN FH, BREZIS M, SIIvA P, ROSEN 5: Physiological and clinicalimplications of medullary hypoxia. .4rtif Organs 11:463—467, 1987

10. PFALLI:R W, RITrINGER M: Quantitative morphology of the ratkidney. IntJ Biochem 12:17—22, 1980

11. VANDEWALLE A, WIRTHENSOIIN G, HEIDRICI-I JIG, GUDER WG:Distribution of hexokinase and phosphoenolpyruvate carboxykinasealong the rabbit nephron. Am J I'hysiol 240:F492—F500, 1981

12. GUDER WG, WAGNER S, WIRTFIENSOHN G: Metabolic fuels along thenephron: Pathways and intracellular mechanisms of interaction. Kid-ney mt 29:41—45, 1986

13. JOHNS 0: Oszillierende Sauerstoffpartialdrueke in der Nierenrinde.Ausdruck einer dureh den Tuhuloglomerulären-Feedback-Meehanis-mus geregeltcn Nephronfunktion? Thesis, Medizinisehe HochschuleHannover, 1994

14. HOLSTEIN-RAI'III.OU NH: Synchronization of proximal intratubularpressure oscillations: Evidence for interaction between nephrons.PJlugers Arch 408:438—443, 1987

15. LEYSSAC PP, BAUMBACH L: An oscillating intratubular pressureresponse to alterations in Henle loop flow in the rat kidney. ActaScand Physiol 117:415—419, 1983

16. SCHUREK HJ, JOhNS 0: Tuhulo-glomerular feedback preventsnephron oxygen deficiency via TAL-segnlents. (abstract) J Am SocNephml 1:603, 1990

17. METZGER H, HARTMANN M, WADOUI-I F: 1he influence of hemor-rhagic hypotension on spinal cord tissue oxygen tension. Adi' Exp MedBiol 200:223—232, 1985

18. EBERT ilL, GLEADLE JM, O'ROURKE JF, BARTI,EFF SM, PoUI:roN J,RATCLIFFE PJ: Isoenzyme specific regulation of genes involved inenergy metabolism by hypoxia: Similarities with the regulation oferythropoietin. Biochem J 3 13:809—814, 1996

19. SCHNLRMANN J, BRIGUS JP: Function of the juxtaglomerular appara-tus. Control of glomerular hemodynamics and renin Secretion (chapt35), in The Kidney, Physioloxy and Pathophysiolory (2nd cd), edited bySELDIN DW, GIEBISCH U, New York, Raven Press, 1992, pp 1249—1289

20. KIUz W: A periarterial pathway for intrarenal distribution of renin.Kidney lot 31:51—56, 1987

21. LEICHTWEISS HP, LUBBERS DW, WEISS C, BAUMGAR1 L H, RLSUHKEW: The oxygen supply of the rat kidney: Measurement of intrarenalP°2 Pflugers Arch 309:328—349, 1969

22. BACHMANN 5, LE HIR M, ECKARDT KU: Co-localization of erythro-poietin mRNA and ecto-5'-nucleotidase immunoreactivity in peritu-hular cells of rat renal cortex indicates that fibroblasts produceerythropoietin. J Histochem (ytochem 41:335—341, 1993

23. KACZMARCZYK U, REINHAROT HW: Arterial blood gas tensions andacid-base status of Wistar rats during thiopental and halothaneanesthesia. Lab Anim Sci 25:184—190, 1975

24. KARL.SEN FM, HOLSTEIN-RATHLOU NH, LEYSSAC PP: A re-evaluationof the determinants of glomerular filtration rate. Acta Phy.siol Scand155:335—350, 1995

25. SCIIUREK HJ: Die Nierenmarkhypoxie: EinSchlüsscl zum Verstàndnisdes akuten Nierenversagens? KIm Wochenschr 66:828-835, 1988

26. SCI-IUREK HJ, JOST U, BErRAM H, BAUMGARTI. 1-I: Preglomerularcortical oxygen diffusion shunt: A prerequisite for effective erythro-poietin regulation? in Erylhropoietin: From Molecular Structure toClinical Application (vol 76), edited by BALDAMUS CA, SCIGALLA P,WIECZ0REK L, KOCH KM, Basel, Contrib Nephrol, Karger-Verlag,1989, pp 57—66

27. FRENKEL R: Control of reduced diphosphopyridine nucleotide oscil-lations in beef heart extracts. I. Effects of modifiers of phosphofruc-tokinase activity. Arch Biochem Biophys 125:151—156, 1968

28. FIESS B, B0Iri-ux A: Oscillatory phenomena in biochemistry. Ann RevBiochem 40:237—258, 1971

29. YIP KP, HOLSTEIN-RATHLOU NH, MARSH DJ: Dynamics of TGF-initiated nephron-nephron interactions in normotensive rats andSHR. Am i Physiol 262:F980—988, 1992

30. BASTIN J, CAMBON N, THOMPSON M, LOWRY OH, BURCH H: Changein energy reserves in different segments of the nephron during briefischemia. Kidney mt 31:1239—1247, 1987