THE JOURNAL OF BIOLOQICAL CHEMISTRY Vol. 244, No. 9, Issue of May 10, PP. 2317-2324, 1969

Printed in U.S.A.

Flavin and Pyridine Nucleotide Oxidation-Reduction Changes in Perfused Rat Liver

I. ANOXIA AND SUBCELLULAR LOCALIZATION OF FLUORESCENT FLAVOPROTEINS”

(Received for publication, December 16, 1968)

ROLAND SCHOLZ, RONALD G. THURMAN,$ JOHN R. WILLIAMSON,$ BRITTON CHANCE, AND

THEODOR BUTCHERY

From the Johnson Research Foundation, University of Pennsylvania, Philadelphia, Pennsylvania 19104

SUMMARY

Flavin and pyridine nucleotide fluorescence was continu- ously monitored from the surface of the hemoglobin-free, perfused rat liver with the use of a new double fluorometer.

In response to anoxia or after addition of inhibitors of the respiratory chain, the fluorescence intensity excited at 436 rnk decreased, while that excited at 366 rnb increased, representing reduction of flavin and pyridine nucleotides, respectively. Subsequent reoxidation of pyridine nucleo- tides in the cytosol by the addition of pyruvate was not ac- companied by a change in flavin fluorescence. Maximal flavin reduction could be achieved by complete inhibition of the respiratory chain in the presence of oxygen. Titration with acetoacetate showed that mitochondrial pyridine nucleotides were in equilibrium with the fluorescent flavo- proteins. Microsomal oxidation-reduction reactions ap- peared to be reflected by small changes in pyriciine nucleoticle fluorescence, but not by changes of flavin fluorescence. It is concluded, therefore, that the flavin signal is predominantly due to mitochondrial flavoproteins with approximately equal contributions from flavin pools located on both the substrate and the oxygen side of the rotenone block. The continuous monitoring of flavin and pyridine nucleotide fluorescence as an experimental approach for studies of compartmentation of oxidation-reduction systems is discussed.

The study of metabolic control mechanisms in highly complex systems such as intact tissues has been limited in the past by the lack of suitable methods permitting both cont.inuous and direct measurements. Usually, insight into cellular events is

* This study was supported by Grant GM-12202-05 and by a Grant-in-Aid from the American Heart Association.

1 National Institutes of Health Postdoctoral Fellow. 5 Established Investigator of the American Heart Association. 7 Permanent address, Physiologisch-Chemisches Institut, Uni-

versity of Munich, Munich, Germany.

obtained by static analyses of tissue contents as well as by dynamic analyses of the extracellular fluid. Optical methods have the distinct advantage of monitoring cellular parameters without disruption of the tissue (1). Their application is, however, limited in the presence of hemoglobin because of its sbrong absorption band in the range of visible light (24). One useful optical parameter is the tissue fluorescence excited at distinct wave lengths, introduced in 1959 by Chance and Jijbsis (5). Subsequently, a close correlation between a blue fluores- cence excited at 366 rnp and the content of reduced pyridine nucleotides of various tissues has been clearly established (6-8). Furthermore, studies of absorbance and fluorescence changes in isolated mitochondria indicate that a yellow fluorescence excited at 436 rnp is due to certain oxidized flavoproteins (9).l

Recently, Chance, Mayer, and Legallia$ developed a double fluorometer for the simultaneous measurement of flavin and pyridine nucleotide fluorescence in suspensions of isolated mito- chondria (10). The purpose of the present paper is to describe the application of this new apparatus to the isolated, hemoglobin- free, perfused rat liver. It was previously shown that changes in fluorescence intensity excited on the liver surface by 366 rnp light represent oxidation-reduction changes in both mitochon- drial and estramitochondrial NAD systems (8, 11). In this study, however, we will present evidence that a change in fluores- cence intensity excited at 436 mp is predominantly due to mito-

chondrial flavoproteins. Simultaneous measurement of the fluorescence changes due to flavin and pyridine nucleotides, therefore, allows the study of compartmentation of oxidation- reduction systems. Preliminary results have been reported previously (12, 13).

EXPERIMENTAL PROCEDURE

Methods

Hemoglobin-free Perfusion of Isolated Rat Liver-The perfusion system described by Scholz and Biicher (7) and by Scholz (8)

1 RAMIREZ, J., AND VEGA, J., RBsumenes 7th Cong. Nat. Cienc. Fisiol., 1964, p. 69.

2 B. Chance, D. Mayer, and V. Legallais, manuscript in preparation.

2317

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

is used with slight modifications. The perfusion fluid consists elsewhere (15). The advantage of the method is that an ische- of 100 ml of Krebs-Henseleit buffer, pH 7.4 (14), containing 4 mia or even hypoxia is totally avoided, since the perfusion is g of bovine serum albumin per 100 ml (Fraction V, Sigma; twice started in situ by inserting a polyethylene cannula into the portal dialyzed). This solution is saturated in a temperature regulated vein with perfusion fluid running. (37’) disc oxygenator, usually with mixtures of oxygen and car- Oxygen concentration in the effluent perfusion fluid is mon- bon dioxide (95 :5), or with mixtures of nitrogen and carbon itored polarographically with a modified Clark type platinum dioxide (95 :5). After the gas supply is changed, 90% saturation electrode.8 occurs in less than 1 min, indicating the high efficiency of the Surface Fluoromctq-The mechanical and electronic details of oxygenator. After passing through a gauze filter placed in a the double fluorometer are described elsewhere.2 The apparatus bubble trap, the saturated perfusion fluid is pumped through (Fig. 1) consists of a high intensity mercury arc lamp, reference the liver via a cannula inserted in the portal vein. It leaves the and measure photomultipliers, combinations of primary and liver via a cannula inserted in the vena cava, and passes an oxy- secondary filters mounted on a rapidly rotating disc, and neces- gen electrode before returning to the oxygenator. With higher sary electronic components. The liver is alternately illuminated perfusion rates than previously described (S), suflicient oxygena- with pulses of 366 and 436 rnp light, the rotation of the disc tion of the peripheral hepatic tissue is maintained without the being used as a time sharing device. The emitted fluorescence use of back pressure. Thus, the problems of swelling and dimin- is detected by photomultipliers after passing through secondary ished bile production are avoided. filters having transmission maxima at 450 and 570 rnp, respec-

Male albino rats, Holtzman strain, 200 to 220 g, are deprived tively. of food for 24 to 30 hours prior to anesthesia with pentobarbital Solutions of reduced pyridine nucleotides have fluorescence (50 mg per kg). Details of the surgical procedure are described maxima at about 470 w when excited with 366 rnp light (16,

17). Moreover, this maximum is shifted to shorter wave lengths after binding to certain enzymes (18). We detect a fluorescence maximum between 440 and 450 rnl.c when the liver surface is illuminated with 366 rnp light (Fig. 2), suggesting that pyridine nucleotides in the intact cell are predominantly bound to pro- tein. Although oxidized flavin nucleotides emit a yellow fluores- cence when excited with 436 rnE.1 light (19), most isolated flavo- proteins are not fluorescent (20). In suspensions of isolated mitochondria, however, changes of fluorescence with a maximum at about 570 rnp are closely correlated with absorption changes of flavoproteins (9, 21). This fluorescence maximum is also ob- served under similar conditions in the perfused liver.

231.8 Oxtiation-Reduction Changes in Perfused Liver. I Vol. 244, No. 9

Materials

Reagent grade chemicals were obtained from Sigma, Calbio- them, Aldrich, or K and K Chemicals. Acetoacetate was pre- pared from the methyl ester (22). Rotenone was dissolved in

FIG. 1. Double fluorometer and perfusion apparatus. PM, dimethylsulfoxide (100 mg per ml), antimycin A in acetone (50 photomultiplier. mg per ml).

436 RESULTS

. C3/cle of A&-Distinct and reversible intracellular oxida- tion-reduction changes are obtained when liver is subjected to a

I cycle of anoxia (7). After oxygen withdrawal, the intensity of the fluorescence excited at 436 mp decreases, while that excited

\ ‘1

at 366 ml.c increases, representing a reduction of both flavin and \

\/I I i \ pyridine nucleotides. The changes roughly index the maximal

/I oxidation-reduction shift possible from the aerobic steady state I i I I ,

300 350 400 450 500 550 600 650 level. They are reversible after restoration of the oxygen supply, Wavelength [ mp] indicating a reoxidation of the coenzymes. Prom a comparison

Secondary Fdter H + of the kinetic changes in the fluorescence excited at 366 rnl.c with

tissue analyses of reduced pyridine nucleotides, it was concluded FIG. 2. Excitation and emission spectra recorded from the

surface of hemoglobin-free, perfused rat liver. The surface was that these changes predominantly represent changes in tissue

illuminated with light of a mercury arc lamp at 366 rnp (-) or NADH content with only a minor contribution of NADPH (7,8).

436 rnrr (- - -). The reflected excitation light and the correspond- So far, such a direct correlation between fluorescence changes ing fluorescence were recorded by a photomultiplier after passing and tissue content is not available for flavoproteins. However, a movable diffraction grating. Broad fluorescence maxima were observed at about 446 nm (-) and 665 rnr (- - -). The band *The perfusion apparatus and oxygen electrodes were con- widths of secondary filters with transmission maxima at 450 and structed in the workshop of the Institute for Physiological Chem- 570 mr, as used in the double fluorometer, are indicated below the istry, University of Munich. The authors thank Dr. B. Brauser figure. _ _ _ . . _ _ . _

for the kind gift of miniature oxygen electrodes.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Issue of May 10, 1969 Scholx, Thurman, Williamson, Chance, and B&her 2319

excitation and emission spectra of liver slices clearly identify this component as flavoprotein, particularly at low temperatures.4 In the following, we will describe a fluorescence increase excited at 366 rnp as pyridine nucleotide reduction, and a fluorescence decrease excited at 436 rnp as flavin reduction.

In the initial phase of the anoxic cycle, a small increase of the pyridine nucleotide fluorescence is observed, while no change in flavin fluorescence is recorded (Figs. 3 and 4, Phase I). Twenty seconds after the onset of anoxia, a rapid flavin reduction occurs, which is approximately paralleled by the pyridine nucleotide reduction (Phase II). Thereafter, the pyridine nucleotide re- duction continues at a slower rate, while considerably less change is observed in the flavin trace (Phase III).

The reduction of cytochromes a and c reaches about 50% before the onset of the rapid pyridine nucleotide reduction occurs (23). Furthermore, a decrease in phosphate potential precedes the rapid increase in tissue NADH (7). It is concluded, there- fore, that cytochromes, phosphate potential, and flavoproteins react in anoxia prior to the major pool of pyridine nucleotides. However, a small component of the pyridine nucleotide pool, detectable by tissue analysis (8) and fluorescence (Phase I), becomes reduced before an apparent reaction of the respiratory chain.

When oxygen is restored to the perfusion fluid, a rapid oxida- tion of flavoproteins is followed by an oxidation of pyridine nucleotides (Phase IV). The original oxidation-reduction state of pyridine nucleotides is slowly re-established. The kinetics of flavoprotein reoxidation shows an “overshoot” phenomenon. It is suggested that immediately after oxygen restoration, elec- tron flux through the respiratory chain is maximal, owing to the excess ADP produced during anoxia, i.e. State 5 + 3 transition (24). Thus, the highly oxidized state of the oxidation-reduction carriers which is observed in State 3 of isolated mitochondria could explain this overshoot. A similar reaction of the mito- chondrial pyridine nucleotides is apparently masked by the extramitochondrial pool. Concomitant with the completion of ADP phosphorylation, a State 3 -+ 4 transition occurs, which is accompanied by a partial reduction of the electron carriers.

Subcellular Localization of Fluorescent Flavoproteins--In pre- vious studies with perfused liver, it was shown that both mito- chondrial and extramitochondrial NADH pools are detectable by surface fluorometry (8, 11). Since the present paper is the first detailed report of flavin fluorescence continuously monitored in intact tissue, we must resolve which flavins are being measured. It is known that flavoproteins of isolated mitochondria have different quantum efficiencies (21). Also, fluorescent flavins have recently been detected in the microsomal fraction of rat liver.’ However, a quantitative comparison of flavin fluores- cence in isolated subcellular fractions with that of intact tissue is not possible, owing to their different fluorescence enhancement and quenching properties. The subcellular localization of fluorescent flavoproteins in perfused liver, therefore, must be studied indirectly by comparing multiple oxidation-reduction changes under different metabolic conditions.

This problem can be approached by using inhibitors known to interact at specific points in the respiratory chain (Fig. 5). Rote- none and Amytal inhibit between two flavoproteins, F,D16 and

4 B. Chance and G. Van Rossum, unpublished observations. 5 I. Hassinen and R. W. Estabrook, unpublished observation. 6 See the legend to Fig. 5 for an explanation of the abbreviations

F,DI and F,D2.

CYCLE OF ANOXIA IN PERFUSED RAT LIVER

Fluorescence Increase

366+450mp

I 4”/ Change

t ‘Fluorescence

Decrease 4364570mp

mM O2

0.3

0.2

0.1

0

FIG. 3. Flavin and pyridine nucleotide oxidation-reduction changes during a cycle of anoxia in the perfused liver. An up- ward deflection reflects a reduction of pyridine nucleotides (upper trace, increase of fluorescence intensity excited at 366 rnp) and flavo- proteins (middle trace, decrease of fluorescence intensity excited at 436 mp). The lower trace represents the oxygen concentration in the effluent perfusion fluid, measured continuously with a platinum electrode. At the points indicated, the gas supply to the oxygenator was changed from OZ-CO2 (95:5) to N&O% (95:5), and vice versa. A time lag of 15 to 20 set is due to both gas exchange and dead space in the perfusion system before anoxia or oxygen restoration occurs in the liver tissue. Different phases (I to V) of the anoxic cycle (7) are indicated by dotted lines.

yi lull IC

ucleotide eduction

L

eduction t

t -It+ N*

IOsec

FIG. 4. Oxidation-reduction changes during a normoxia-anoxia transition (7) in the perfused liver: oscilloscope recording of flavin (lower trace) and pyridine nucleotide (upper trace) fluorescence. An upward deflection represents reduction of these coenzymes. One vertical scale division represents a 2y0 change in fluorescence intensity.

F,Dz (NADH dehydrogenase (EC 1.6.99.3)), which are char- acterized by different fluorescence yields and oxidation-reduction potentials (21, 26). F,D1 is thought to be coupled with li- poamide dehydrogenase, and is reduced by NADH, cr-keto-

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2320 Oxidation-Reduction Changes in Perfused Liver. I Vol. 244, No. 9

glutarate, or pyruvate, whereas F,D2 transfers electrons to the cytochrome chain (10, 27). F,DQ is coupled with succinate de- hydrogenase, and is reduced in the rotenone-blocked system only by succinate, n-glycerophosphate, or fatty acids. Although antimycin A and sulfide inhibit at different points, they both block the transfer of electrons from flavoproteins to oxygen. Thus, anoxia and inhibitors of electron flux in the cytochrome chain should cause the same degree of reduction of mitochondrial flavoproteins. Furthermore, the problem of subcellular localiaa- tion can be approached by compensating oxidation-reduction changes by the addition of pyruvate or acetoacetate, substrates known to react with dehydrogenases in different compartments.

Rotenone-Shortly after the addition of rotenone, a rapid flavin reduction is observed, followed after a time lag of about 2 set by a reduction of pyridine nucleotides (Fig. 6). Maximal responses are reached in 2 min. The half-times of both reac- tions are about 30 sec. Occasionally, a slight reoxidation of pyridine nucleotides occurs, reaching a constant level after about 5 min (cf. Fig. 10, below). Concomitant with the oxidation- reduction changes, the oxygen consumption decreases about 50%

Amytal Rotenone Antimycin A Sulfide

NADH- w Cytochromes

cr-Ketoglutarote Succinate Pyruvate

FIG. 5. Scheme for the mitochondrial flavin chain (25). F,D, NADH dehydrogenase (EC 1.6.99.3) ; F,L, lipoamide dehydro- genase (EC 1.6.4.3) ; FpS, succinate dehydrogenase (EC 1.3.99.1). Rotenone or Amytal separates F,DI from F,Dz.

Flavin Reduction t

Pyridine Nucleotide Reduction t

t -4 I+-IIsec Rotenone 25,uM

FIG. 6. Fluorescence changes following the addition of rotenone to the perfusion fluid: oscilloscope recording of flavin (upper trace) and pyridine nucleotide (lower trace) fluorescence. An upward deflection represents reduction. Rotenone (1 mg) was dissolved in dimethylsulfoxide (10 ~1) and added to a sample of perfusion fluid with rapid stirring. Since precipitation of rote- none could not be avoided, the final concentration should be less than 25 PM. One vertical scale division represents 4y0 change in fluorescence intensity.

I 1 W

El%@ - 4%

I ! I I 1 1 4rpin+--+/

Pyridine Nucleotide t

Reduction

Flavin Reduction t

Fluorescence Change

FIG. 7. Effect of pyruvate on flavin and pyridine nucleotide reduction in the perfused liver following inhibition by rotenone. Eight minutes after the addition of rotenone, 0.2 pmole of sodium pyruvate was slowly added to the perfusion fluid. Pyruvate, which gives maximal effects with a concentration of 2 mM, reverses 40?& of the increment of pyridine nucleotide fluorescence observed after rotenone addition, if the base-line drift (see left hand side of the trace) is considered. We cannot explain why the flavin reduction in this experiment is slower than that usually observed (see Figs. 6, 9, and 10).

(Fig. 10). Based on the known site of rotenone ir&ibition, the fluorescence changes are expected to represent primarily the reduction of respiratory chain components located on the sub- strate side of the rotenone block, i.e. F,Dr-lipoamide dehydrogen- ase complex and mitochondrial NAD. Extramitochondrial NAD, however, is also expected to be involved, owing to the stimulation of glycolysis following a decreased phosphate po- tential.7

Compensation of Oxidation-Reduction Changes-It can be con- cluded that in anoxia or with rotenone or Amytal, pyruvate reacts predominantly with extramitochondrial NADH, whereas mitochondrial reactions, i.e. pyruvate carboxylation and de- carboxylation, are prevented because of the lack of energy and NAD. Furthermore, the first step in pyruvate decarboxylation appears to be unlikely since lipoamide dehydrogenase would already be reduced, owing to its rapid equilibration with NADH dehydrogenase (25). The reoxidation of pyridine nucleotides reduced by rotenone following the addition of pyruvate (Fig. 7, about 40%) and the diminished reduction during anoxia in the presence of high pyruvate concentrations (Fig. 8) are, therefore, primarily due to an extramitochondrial NADH pool which is in connection with lactate dehydrogenase, as shown previously.’

Flavin reduction is not diminished during anoxia in the pres- ence of high pyruvate concentrations compared to a control in the absence of a hydrogen acceptor for extramitochondrial NADH (Fig. 8). Likewise, less than 10% of the flavoproteins reduced by rotenone can be reoxidized by the addition of pyruvate (Fig. 7). Thus, the reoxidation of extramitochondrial NADH does not affect the oxidation-reduction state of fluorescent flavopro- teins.

7 R. Schols, J. Grunst, F. Schwars, and T. Bticher, manu- script in preparation.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Issue of May 10, 1969 Scholz, Thurman, Williamson, Chance, and Biicher 2321

CYCLE OF ANOXIA

k--ZmM+ B Pyruvate Infusion

A Control

FIG. 8. Flavin and pyridine nucleotide oxidation-reduction changes of the perfused liver during a cycle of anoxia in the presence and in the absence of high pyruvate concentrations. One minute after the gas supply to the oxygenator was switched from Oz-CO2 to NvCO~, pyruvate was infused into the perfusion fluid before it entered the liver, maintaining a concentration of approximately 2 mM. An accumulation of pyruvate could occur after 2 min of infusion, owing to the turnover in the perfusion system (volume, 100 ml; flow rate, 40 ml per min). After 4 min of anoxia, the gas supply to the oxygenator was switched to O&YJOn; the pyruvate infusion was stopped.

Further support is obtained from similar experiments in which acetoacetate is added after rotenone inhibition (Fig. 9). High concentrations of acetoacetate in the perfusion fluid only par- tially reoxidized pyridine nucleotides (about 70%), but reoxi- dize flavoproteins completely. In contrast with pyruvate, aceto- acetate is a hydrogen acceptor for mitochondrial NADH, its main interaction in liver being with /3-hydroxybutyrate dehydrogenase. This experiment indicates that part of the NAD reduced by rotenone is not in equilibrium with this dehydrogenase, i.e. NAD in the cytosol. Conversely, the entire flavoprotein pool reduced by rotenone appears to be in equilibrium with mitochondrial NADH linked to fi-hydroxybutyrate dehydrogenase.

Antimycin A and Suljkk-Another flavin pool containing F,D2, succinate dehydrogenase, and the flavoproteins, involved in fatty acid oxidation remains oxidized in the rotenone-inhibited state. This pool becomes reduced when the electron flux to oxygen is completely stopped by anoxia or by addition of anti- mycin A (Fig. 10). The flavin reduction observed shortly after the addition of antimycin A exceeds that reached in 4 min of anoxia; however, nearly the same level of reduction is reached if the anoxia is continued for a longer time period (not shown in this experiment). The changes in flavin fluorescence caused by rotenone and subsequently by antimycin A are almost equal.

In the presence of rotenone, the reoxidation of pyridine nucleo- tides after anoxia is incomplete. However, the same fluorescence increase is reached with antimycin A in comparable time intervals as with anoxia Furthermore, this pyridine nucleotide reduction continues, while the flavin reduction maintains a maximal level. Assuming that the oxidation-reduction state of flavoproteins reflects that of mitochondrial NAD, the prolonged pyridine nucleotide reduction following antimycin A inhibition indicates the absence of a rapid equilibrium between extra- and intramito- chondrial pools.

The fact that anoxia does not cause a larger flavinreduction than inhibition with antimycin A is confirmed in a similar experiment with sulfide (Fig. 11). Oxygen withdrawal after maximal inhi- bition with sodium sulfide does not further reduce fluorescent

t Pyridine

Nucleotide Reduction

25pM Rotenone

Fluorescence Change

t Flavin

Reduction

FIG. 9. Effect of acetoacetate on flavin and pyridine nucleotide reduction in the perfused liver following inhibition by rotenone. Acetoacetate was infused over two time periods as indicated by the bars to give a final concentration near 6 mM. The turnover time of the perfusion system of 23 min was longer than the first infusion period, accounting for the transient reversal of the aceto- acetate effect.

Pyridine

Nucleotide

Reduction

t

Flavin

Reduction I

L!, I I I I I 72 “2 I I I 1 02

El

In

:fluent

50 60 MinutesOf Perfusion

FIG. 10. Changes of flavin (middle trace) and pyridine nucleo- tide (upper trace) fluorescence of perfused liver during a cycle of anoxia and following the addition of rotenone and antimycin A. Oxygen concentration in the effluent perfusion fluid is shown by the lower truce. The concentration entering the liver was 0.72 mM 0%. Anoxia was performed, and rotenone and antimycin A were added as described in the legend to Fig. 6. One vertical scale division represents 3% change in fluorescence intensity.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2322 Oxidation-Reduction Changes in Perfused Liver. I Vol. 244, No. 9

flavoproteins, indicating. that in the absence of sulfide or anti- mycin A the flavin reduction observed in anoxia is due to mito- chondrial flavoproteins. Unexpectedly, pyridine nucleotide reduction increases rapidly after oxygen withdrawal in the presence of sulfide. Since the egress of reducing equivalents from the mitochondria as well as the generation of NADH in the cytosol should not be accelerated under these conditions, the

Sulfide Anoxia

Pyridine Nucleotide T Reduction

I 4% -K

Flu;;wF;;ce

Flavin Reduction T

bZmM-+ Na,S Infusion

FIG. 11. Effect of anoxia on flavin and pyridine nucleotide fluorescence in the perfused liver following an infusion of sodium sulfide (2 mM final concentration).

I I I I I I I >yridine Nucleotide deduction

%!33&

t’ ’ ’ ’ Flavin Reduction 1 I / I I /

L_IU =IrnM Aminopyrinea

6x$& In Effluent I I I I I I I I t r 1

80 100 110

Minufes Of Perfusion

FFIG. 12. Changes of flavin (middle trace) and pyridine nucleo- tide (upper trace) fluorescence of perfused liver during a cycle of anoxia and after the addition of aminopyrine. Oxygen concen- tration in the effluent perfusion fluid is shown by the lower truce. The concentration entering the liver was 0.7 mM 02. The oxygen consumption of the liver was calculated from the arteriovenous differences, the perfusion rate (25 ml per min), and the wet weight of liver (8.2 g). The transient decrease in effluent oxygen concen- tration after aminopyrine addition reflects a 5% increase in oxygen consumption by the liver.

interaction of reduced pyridine nucleotides with extramitochon- drial oxidases should be considered.

Microsomal Oxidation-Reduction Reactions-It might be ex- pected that extramitochondrial flavoproteins should also be re- duced in anoxia. Cyanide-insensitive respiration was found to be 10 to 20% of the total oxygen uptake of the perfused liver (7, 15), indicating the activity of extramitochondrial oxidases. The negligible contribution of these flavoproteins to the fluores- cence changes in anoxia could be due, first, to a lower quantum efficiency of their fluorescence, or second, to smaller oxidation- reduction changes to that of flavins linked to the respiratory chain. However, if microsomal oxidation-reduction changes are detectable, they should become apparent when substrates for mixed function oxidation, such as aminopyrine, are metabolized. Shortly after the addition of aminopyrine, an increase in oxygen consumption of about 5% and an oxidation of pyridine nucleo- tides occur (Fig. 12). These responses indicate that microsomal flavin-enzymes, such as NADH-NADPH diaphorase (EC 1.6.1.1) and NADPH-cytochrome c reductase, are reacting; but no change in flavin fluorescence is observed. Furthermore, it was recently suggested that ethanol is a substrate for the mixed func- tion oxidase system (28). Usually, ethanol metabolism is re- flected by a reduction of pyridine nucleotides and flavoproteins (12). However, in the presence of pyrazole, a potent inhibitor of alcohol dehydrogenase (29), addition of ethanol has no effect on flavin fluorescence, indicating that microsomal ethanol oxidation does not cause detectable fluorescence changes, if the assumpt,ion is made that pyrazole is without effect on this system. Thus, no evidence is obtained that changes in flavin fluorescence must be attributed to flavoproteins located outside mitochondrial mem- branes.

DISCUSSION

Determination of Oxidation-Reduction States in D@erent Xub- cellular Compartments of Intact Tissue-Studies of oxidation- reduction compartmentation are confronted with the experi- mental problem that tissue analyses do not allow insight into the oxidation-reduction states of different subcellular compartments. Furthermore, these different oxidation-reduction states cannot be maintained during the process of subcellular fractionation at the present time. Dissection of subcellular components from tissues followed by microanalysis has been accomplished in a variety of cells (30, 31). Although this laborious technique has not been applied to problems of oxidation-reduction compartmentat,ion, it seems to be possible. A completely different approach, de- veloped mainly by Biicher and Klingenberg (32), makes the assumption that thermodynamic equilibrium is obtained in dif-

ferent oxidation-reduction systems. Subcellular oxidation- reduction states, therefore, are calculated from the tissue content or extracellular concentrations of substrate couples, the dehydro- genases of which are localized exclusively in one compartment.

Much attention has been devoted to the continuous read-out of metabolic signals from intact tissues (I). With respect to oxida- tion-reduction studies, such signals may be derived from the fluorescence of coenzymes, particularly reduced pyridine nucleo- tides, as shown by Chance and Jobsis for intact sartorius muscle (5). Two general principles are followed. First, the fluores- cence of different cellular organelles can be observed by micro- fluorometry. This method has been successfully applied to the direct read-out of pyridine nucleotide reduction, mainly in perinuclear mitochondria and the clear cytoplasm of single cells

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Issue of May 10, 1969 Xcholx, Thurman, Williamson, Chance, and Biicher 2323

(33). A second approach is the use of fluorescence signals de- rived exclusively from one subcellular compartment. While pyridine nucleotide fluorescence is derived almost equally from both mitochondria and cytosol, flavin fluorescence of liver tis- sue, as shown in these experiments, is derived predominantly from mitochondrial flavoproteins. Since these flavoproteins are thought to be closely equilibrated with mitochondrial NADH, the over-all pyridine nucleotide signal of the tissue may be corrected for the mitochondrial component.

Oxidation-Reduction Changes of Mitochondrial Flavoproteins In this study, several indirect approaches have been used to show that flavin fluorescence in the perfused liver is an indicator for the mitochondrial space. First, the addition of a hydrogen acceptor for extramitochondrial NADH, such as pyruvate, in the rotenone-inhibited or anoxic state causes a considerable reoxida- tion of pyridine nucleotides, but only a negligible reoxidation of flavoproteins. However, a complete reoxidation of flavoproteins reduced by rotenone occurs after the addition of acetoacetate, an acceptor for mitochondrial NADH. This flavin pool, therefore, represents the oxidation-reduction state of mitochondrial, rather than extramitochondrial, NADH.

Moreover, a complete reoxidation of reduced pyridine nucleo- tides by high concentrations of either pyruvate or acetoacetate was not possible under these conditions. This observation agrees with similar experiments, where the reoxidation in anoxia (8) and in the Amytal-inhibited state7 was titrated. Thus, the with- drawal of reducing equivalents from one compartment does not effect a rapid efflux from the other compartment. Since rote- none, Amytal, and anoxia all decrease the energetic state of the liver cell, it is possible that these two compartments fail to int.er- act rapidly because of a lack of energy.

Second, anoxia did not cause a greater flavin reduction than inhibition of the respiratory chain between oxygen and the flavin region. A possible reduction of microsomal flavoproteins in anoxia, therefore, can give only a minor contribution to the lluo- rescence signal, This is also shown under conditions of increased mierosomal activity, in which no changes of flavin fluorescence are detectable. Ultimately, a more accurate analysis may lead to direct measurements of flavoproteins associated with extramito- chondrial processes as well. However, at present, they constitute such a small portion of the total signal in the types of metabolic transitions studied here that they may be neglected.

All of the available experimental evidence, therefore, confirms the conclusion that more than 90% of the flavin fluorescence changes observed from the liver surface is due to mitochondrial flavoproteins. In a transition from the aerobic steady state to a completely reduced state, the flavin pools on either side of the rotenone block give almost equal contributions to the fluores- cence changes, However, less than 30% of the total flavin ex- tractable from isolated liver mitochondria can account for a flavin pool on the substrate side of the rotenone block.* This confirms the observation from mitochondrial experiments that this pool has a higher fluorescence yield than the one on the ,oxygen side (20).

Kinetics of Intracellular Oxidation-Reduction Reactions in TissuesOf importance to the interpretation of experimental data on intracellular reactions and on the modelling of intracellu- lar reactions by computer techniques is information on the speed with which these reactions occur. Kinetics of reactions in sub-

* R. G. Thurman, R. Scholz, J. R. Williamson, I. Hassinen, and B. Chance, manuscript in preparation.

cellular fractions and single cells can be determined by the use of techniques of rapid mixing. But these techniques, such as the flow apparatus (34), are not applicable to intact tissues. In a perfused organ, in which cellular reactions are initiated via the transport of substances by the perfusion fluid, the time difference between the onset in different cells can reach several seconds. The initial phase of fluorescence changes, therefore, is influenced by the statistical distribution of a population of lo6 to lo7 liver cells. Thus, the observed kinetics of the “on” reaction in anoxia or rotenone inhibition with half-times of 15 and 30 set, respec- tively, is much slower (at least one order of magnitude) than in experiments with isolated mitochondria (27).

In the “on” reaction of anoxia, the reduction of cytochromes a and c (23) and of flavoproteins occurs before the reduction of the major portion of pyridine nucleotides. A minor portion, how- ever, becomes reduced before any reaction of the respiratory chain is detectable. This phenomenon is not observed in an aerobic- anaerobic transition of isolated mitochondria or with rotenone inhibition in perfused liver. Thus, the reduction in Phase I of anoxia appears to be independent of the respiratory chain. It may be due to a pool of pyridine nucleotides coupled to an oxy- gen acceptor with a K, for oxygen greater than that of cyto- chrome oxidase, such as oxidases located in peroxisomes (35) or in the smooth endoplasmic reticulum.$

Studies of Compartmentation of Oxidation-Reduction Systems- The subcellular distribution of fluorescent flavoproteins, as shown in this paper, has unique applications to the study of compartmentation of oxidation-reduction systems. Thus, flavin or pyridine nucleotide fluorescence can be used as a continuous monitor for mitochcndrial or for mitochondrial plus cytosolic oxidation-reduction states, respectively.

A decrease of the electron flux in the respiratory chain, ini- tiated either by oxygen withdrawal or by the addition of specific inhibitors, causes primarily a reduction of mitochondrial pyri- dine nucleotides and a fall in the phosphate potential. A conse- quent reduction of extramitochondrial pyridine nucleotides could be due to the following possibilities: (a) an egress of reducing equivalents from the mitochondria, (b) a decreased rate of extra- mitochondrial oxidation of NADH and NADPH, and (c) an increased generation of NADH in the glycolytic chain stimulated by the adenosine phosphate system. The experiments described in this paper give no evidence that a flux of reducing equivalents across mitochondrial membranes is involved. Since the oxida- tion-reduction changes in both cytosol and mitochondria can be separately compensated by specific hydrogen acceptors, it was previously7 concluded that a rapid bidirectional hydrcgen flux does not occur when the respiratory chain is inhibited, i.e. in the absence of energy. Furthermore, evidence exists that, in anoxia, a decreased extramitochondrial oxidation of pyridine nucleo- tides is, in part,, involved. The reduction in Phase I and the increased pyridine nucleotide reduction in anoxia after prior inhi- bition with sulfide may be ascribed to an inhibition of the extra- mitochondrial oxidases. Coupling of extra- and intramitochon- drial oxidation-reduction reactions via the adenosine phosphate system appears to be the important mechanism of cytosolic pyridine nucleotide reduction following a decreased electron flux in the respiratory chain.

Under more nearly physiological conditions, i.e. in the presence of energy sources, a bidirectional flux of reducing equivalents

9 D. Y. Cooper and V. Ullrich, unpublished observations.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2324 Oxidation-Reduction Changes in Perfused Liver. I Vol. 244, No. 9

across the mitochondrial membranes is observed to occur. Dur- 14. KREBS, H. A., AND HENSELEIT, K., Hoppe-Seyler’s 2. Physiol.

ing ethanol or lactate oxidation, this flux is directed into the Chem., 210, 33 (1932).

mitochondria, whereas during gluconeogenesis or fatty acid 15. SCHWARZ, F., Thesis, Medical Faculty, University of Munich,

oxidation the flux is in the opposite direction. Germany, 1967.

The simultaneous 16. WARBURG, 0.. Wasserstoff tibertraaende Fermente. Saenaer, __ monitoring of flavin and pyridine nucleotide fluorescence and Berlin, 1948,‘~. 21. - ”

I I

oxygen consumption allows the study of speed and extent in the 17. THEORELL, H., AND MCKINLEY-MCKEE, J. S., Acta &em.

intracellular transfer of reducing equivalents, as well as the rela- Scand., 16, 1811 (1961).

18. CHANCE, B., AND BALTSCHEFFSKY, H., J. Biol. Chem., 233, tive involvement of different transport shuttle mechanisms. 736 (1958):

19. THEORELL. H.. Biochem. 2.. 278.263 (1935).

REFERENCES

1. CHANCE, B., Harvey Lect., 49, 145 (1955). 2. CHANCE, B., COHEN, P., J~~BSIS, F., AND SCHOENER, B., Science,

137, 499 (1962.) 3. CHANCE, B., AND &XOENER, B., Biochem. Z., 341, 340 (1965). 4. SCHNITGER, H., SCHOLZ, R., B~CHER, T., AND LUBBERS, D. W.,

Biochem. Z., 341,334 (1965). 5. CHANCE, B., AND J~~BSIS, F., Nature, 184,195 (1959). 6. CHANCE, B., WILLIAMSON, J. R., JAMIESON, D., AND SCHOENER,

B., Biochem. Z., 341,357 (1965). 7. SCHOLZ, R., AND BUTCHER, T., in B. CHANCE, R. W. ESTABROOK,

AND J. R. WILLIAMSON (Editors), Control of enerau metabo- lism, Academic Press, New York; 1965, p. 393. “_

8. SCHOLZ, R., in W. STAIB AND R. SCHOLZ (Editors), Sto$wechseZ der perfundierten Leber, Springer-Verlag, Berlin, 1968, p. 25.

9. CHANCE, B., AND SCHOENER, B., in E. C. SLATER (Editor), Flavins and flavoproteins, Academic Press, New York, 1966, p. 510.

10. LEE, I. Y., AND CHANCE, B., Biochem. Biophys. Res. Commun., 32, 547 (1968).

11. WILLIAMSON, J. R., BROWNING, E. T., AND OLSON, M. S., Advan. Enzvme Reaul.. 6. 67 (1968).

12. SCHOLZ, R., A&D THURMAN; R. b., Fkd. Proc., 27, 462 (1968). 13. WILLIAMSON, J. R., SCHOLZ, R., THURMAN, R. G., AND CHANCE,

B., in E. QUAGLIARRIELLO, E. C. SLATER, S. PAPA, AND J. M. TAGER (Editors), Energy level and metabolic control in mito-

chondria, Adriatica Editrice, Bari, Italy, in press.

20.

21.

22. 23.

24.

25. 26.

27. 28.

29.

30.

31.

32.

33.

34. 35.

KING, T., in E’. C. SLATER iEditor), Plav&s and llavoproteins, A&de&c Press, New York, 1966; p. 522. - -

CHANCE. B.. ERNSTER. L.. GARLAND. P.. LEE. C. P.. LIGHT. P. A.,‘~H&sHI, T., GAG&, C. I., A&D %ONG; D., P;oc. Nat: Acad. Sci. U. S. A., 67, 1498 (1967).

HALL, L. M., Anal. Biochem., 3, 75 (1962). LUBBERS. D. W.. KESSLER. M.. SCHOLZ. R.. AND BUTCHER, T..

Biochek Z., 341, 346 (1965). ’ ’ ’ ,

CHANCE, B., AND WILLIAMS, G. R., J. Biol. Chem., 217, 409 (1955).

CHANCE, B., Fed. Proc., 27, 467 (1968). HASSINEN, I., AND CHANCE, B., Biochem. Biophys. Res. Com-

mun., 31, 895 (1968). LEE, I. Y., Fed. Proc., 27, 464 (1968). RUBIN, E., HUTTERER, F., AND LIEBER, C. S., Science, 169,

1469 (1968). THEORELL. H., AND YONETANI, T.. Biochem. 2.. 338, 537

(1963). ’ HYDEN, H., in H. HYDEN (Editor), The neuron, Vol. I, Ameri-

can Elsevier Publishing Company Inc., New York, 1967, p. 180.

LOWRY, 0. H., ROBERTS, N. R., SCHULZ, D. W., CLOW, J. E., AND CLARK. J. R.. J. Biol. Chem.. 236. 2813 (1961).

B~CHER, T., XND ~LINGENBERG, G., Anger. khek., 70, 552 (1958).

CHANCE, B., KOHEN, E., KOHEN, C., AND LEGALLAIS, V., Advan. Enzyme Regul., 6, 3 (1967).

CHANCE, B., Nobel Symp., 6, 437 (1968). DE DUVE, C., AND BAUDHUIN, P., Physiol. Rev., 46,323 (1966).

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Theodor BücherRoland Scholz, Ronald G. Thurman, John R. Williamson, Britton Chance and

FLUORESCENT FLAVOPROTEINSLiver: I. ANOXIA AND SUBCELLULAR LOCALIZATION OF

Flavin and Pyridine Nucleotide Oxidation-Reduction Changes in Perfused Rat

1969, 244:2317-2324.J. Biol. Chem. 

  http://www.jbc.org/content/244/9/2317Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/244/9/2317.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from