THE JOURNAL OF B~LOGICAL CHEMISTRY Vol.245, No. 18,Issue of September 25, pp. 4724-4731, 1970

Printed in U.S.A.

The Metabolism of Rat Brain Mitochondria

PREPARATION AND CHARACTERIZATION*

(Received for publication, March 10, 1970)

JOHN B. CLARKE AND WILLIAM J. NICKLASS

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

SUMMARY

Since most previous studies on brain mitochondria have used a relatively crude preparation, a method was developed for preparing a purified mitochondrial fraction from rat brain cerebral cortex utilizing a rapid simple discontinuous Ficoll density gradient procedure. This mitochondrial preparation was shown by electron microscopy and enzymatic assay to be (a) free from contamination by synaptosomes or other mem- branous fragments and (b) predominantly of neuroglial origin. These mitochondria exhibited good respiratory control and respiratory rates with various substrates (pyruvate > suc- cinate > glutamate > citrate, acetylcarnitine, isocitrate, cu-ketoglutarate, a-glycerophosphate), comparable with other mammalian mitochondria. Pyruvate plus malate was oxi- dized at rates higher than any of the other substrates tested and consistently showed respiratory control ratios greater than seven. The mitochondria were further characterized with respect to nicotinamide and adenine nucleotide con- tent, and cytochrome complement and content. The ratios of cytochrome b:q: C:Q:CQ were 0.6:O.g: 1.7: 1:0.6, similar to those observed in other mammalian mitochondria. As has been noted previously with brain preparations, increasing the K+ in the incubation medium caused an increase in the oxygen uptake with all the above substrates (except or- glycerophosphate). This effect was studied in more detail with pyruvate plus malate as substrate. Half-maximal respiration rates were obtained with 10 to 20 m.M K+ in both ADP and valinomycin-stimulated mitochondria. The effect of K+ was not directly associated with an activation of elec- tron transport per se, but rather with an increase in the availability of substrate for oxidation.

The literature is replete with reports regarding the prepara,tion of mammalian brain mitochondria, e.g. JGbsis (2) cites 19 and Ozawa et al. (3) cite 26 different preparative procedures, but the resulting data have often been contradictory, misleading, or in- complete. Before engaging in an extensive study of the metab-

* This work was supported by Grant GM 12202 from the United States Public Health Service. A preliminary report of this work was presented (1).

$ Medical Research Council Traveling Fellow on leave of ab- sence from Biochemistry Department, St. Bartholomew’s Hos-

olism of brain mitochondria, therefore, the development of a method for the routine preparation of relatively uncontaminated, metabolically active brain mitochondria was of primary impor- tance.

Problems associated with the preparation of brain mitochon- dria, as compared with other tisssues, are often the result of the high lipid content of the brain. This causes the release of mem- branous segments on homogenization which undergo spontaneous vesiculation, leading to the formation of, among other artifacts, synaptosomes (4). There is little doubt that the crude mito- chondrial preparations used by some workers (e.g. 2, 3, 5, and 6) are contaminated with such material, since this is the starting point for the preparation of synaptosomes (4, 7). This may be responsible for the misleading reports of mitochondrial function.

The mitochondrial preparation described here utilizes a density gradient separation procedure with Ficoll which is specifically de- signed to remove synaptosomes and other nonmitochondrial ma- terial. The method has been developed from existing procedures for the preparation of mitochondria and synaptosomes, particu- larly those of Stahl et al. (8) and Basford (9), Kurokawa, Saka- moto, and Kato (lo), Moore (II), and Salganicoff.’

From the results it may be seen that this preparation exhibits good respiratory quotients and control with no requirement for bovine serum albumin (cf. 7, 8) toget’her with properties which suggest that it is exhibiting a purely mitochondrial function, free from synaptosomal contamination.

METHODS

Preparation of Mitochondria-XIale Sprague-Dawley rats (weight 160 to 180 g) fed ad lib&m on laboratory chow were de- capitated and the cerebral hemispheres rapidly removed into ice-cold isolation medium (0.25 M sucrose-10 mM Tris-0.5 mM K+-EDTA, pH 7.4). The tissue was chopped finely with scis- sors while being washed frequently with ice-cold isolation me- dium. The material from eight rats was placed in a Dounce ho- magenizer (Blaessig Glass, Rochester, New York 14009) together with 40 ml of cold isolation medium, and manually homogenized by eight up and down strokes with a glass pestle (total clearance

pita1 Medical College, University of London, Charterhouse Square, London, E. C. 1, England.

5 National Institutes of Health Postdoctoral Fellow, 1968 to 1970. Present address, Department of Neurology, College of Physicians and Surgeons of Columbia University, New York, New York 10032. To whom inquiries regarding this paper should be sent.

1 L. Salganicoff, private communication.

4724

by guest on Novem

ber 17, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of September 25, 1970 J. B. Clark and W. J. Nicklas 4725

0.002 inches). A further 20 ml of ice-cold isolation medium were added and the total homogenate centrifuged at 2’ for 3 min at 2,000 x g. The supernatant from this spin was then centri- fuged for 8 min at 12,500 x g. The crude mitochondrial pellet was resuspended to a final volume of 10 ml in a 3 y0 Ficoll medium (3 y0 Ficoll-0.12 M mannitol-0.03 M sucrose-25 PM K+-EDTA, pH 7.4). This suspension was carefully layered onto 20 ml of a 6% Ficoll medium (6% Ficoll-0.24 M mannitol-0.06 M sucrose-50 pM

K+-EDTA, pH 7.4) and centrifuged for 30 min at 11,500 x g. The supernatant from this spin was decanted and the slight f luffy layer removed from the pellet. The mitochondrial pellet was resuspended in isolation medium and recentrifuged for 10 min at 12,500 x g. The mitochondria were made up to a con- centration of 20 mg of protein per ml in the isolation medium. The average yield per rat brain was between 3 and 4 mg of mito- chondrial protein. Synaptosomes were prepared essentially according to the method of Bradford (12).

Incubation ConditionsOxygen uptakes were measured polaro- graphically with a Clark-type microelectrode (Yellow Springs Instrument Company, Yellow Springs, Ohio). The media used routinely contained either 5 or 150 InM K+ and consisted of 5 mM K+ medium: 225 mM mannitol-75 mM sucrose-5 mM Tris phosphate, pH 7.2-10 mM Tris-Cl, pH 7.4-0.05 mM EDTA-5 mM KCl; all adjusted to pH 7.4 with 2 M Tris; 150 InM K+ me- dium: 5 mM Tris phosphate, pH 7.2-10 mM Tris-Cl, pH 7.4-0.05 mM EDTA-150 mM KCl; all adjusted to pH 7.4 with 2 M Tris. In those cases where the K+ concentration was varied the man- nitol-sucrose concentrations were adjusted to maintain the same osmolarity. The incubation temperature in all cases was 28”.

Fluorometric Procedures-Pyridine nucleotide reduction was followed with a modified Eppendorf fluorometer (13) and in some cases, samples were removed for pyridine and adenine nu- cleotide estimation (14).

Cytochrome Measurements-Low temperature spectra of the cytochromes of mitochondria and synaptosomes were obtained with a split beam scanning spectrophotometer (15, 16). Samples were prepared in cells with a path length of 2 mm by the trapped steady state technique (16). Total cytochrome content was estimated from both the low temperature spectra with the ap- propriate intensification factors (16) and by the use of the Aminco-Chance dual wave length spectrophotometer. The wave lengths used in the latter method were derived from room temperature split beam spectra and the extinction coefficient for each cytochrome was taken from the work of Wilson and Epel

(17). Lactate dehydrogenase activity in subcellular fractions was

estimated by following NADH oxidation at 340 nm in a medium containing 50 rnh< potassium phosphate, pH 7.4, 1 mM pyruvate, 0.2 mM NADH, and 0.5% Triton X100. Succinate dehydrogen- ase was measured by the method of Pennington (18). Proteins were measured by the biuret method (19).

Chemicals-Ficoll was obtained from Pharmacia, Uppsala, Sweden, and purified by dialysis before use. Monofluoroxalo- acetic acid was a gift of Dr. E. Kun to Dr. J. R. Williamson. Pyruvic acid was twice distilled under vacuum and stored at 2”. 911 other substrates were commercial preparations of the highest purity available. Enzymes used for fluorometric analyses of pyridine and adenine nucleotides were purchased from Boeh- ringer Mannheim.

Electron ,Vicroscopy-This was kindly performed by Dr. Caro- line Damsky of the Biology Department, University of Pennsyl-

vania. The mitochondria were pelleted in 0.15 M phosphate buffer and fixed in 3% glutaraldehyde. After rinsing, postfixa- tion was carried out in 2 y0 osmium tetroxide. The preparations were then dehydrated in alcohol, embedded in Epon and sec- tioned with a Porter-Blum MT-2 ultramicrotome. Staining was carried out with uranyl acetate in 50% alcohol and lead citrate. The sections were examined with an AEI EMGB mi- croscope (Associated Electric Industries).

RESULTS

Electron micrographs of Fraction G (Table I) indicated a prep- aration from rat brain which was predominantly mitochondrial and was contaminated only slightly with other subcellular struc- tures and nonspecific membranous material (Fig. 1). The few intact synaptosomes that were present contained mitochondria much different from the bulk of the preparation which suggested that this preparation of brain mitochondria was mainly of non- nerve ending origin.

Further confirmation of the low synaptosomal contamination of this preparation is evident from Table I where the succinate and lactate dehydrogenase activities of each of the fractions of the separation procedure are reported. The crude mitochondrial fraction (Fraction D) possessed a succinate to lactate dehydrogen- ase ratio only slightly higher than the original homogenate (3.8 --f 5.6). Since succinate dehydrogenase may be considered as a mitochondrial ma)rker and lactate dehydrogenase as a cyto- plasmic marker, this fraction (D) was still substantially contami- nated with cytoplasmic inclusions. These inclusions were al- most certainly in the form of synaptosomes since such a crude fraction forms the starting point for a synaptosomal preparation (4). It is also worth noting that this fraction often has been used as the final mitochondrial preparation by other workers (e.g. 2,3,5, and 6). However, after passing such a fraction through a discontinuous Ficoll gradient a 6-fold increase in the succinate to lactate dehydrogenase ratio occurred together with the removal of almost 75% of the protein. A further washing removed more protein and increased the succinate to lactate dehydrogenase ratio still further; the final mitochondrial pellet (G) had a ratio of succinate to lactate dehydrogenase 1 l-fold greater than the origi- nal homogenate. The yield, in terms of mitochondrial protein, was on an average 3 to 4 mg of protein per rat brain.

As this preparation of brain mitochondria had a large propor- tion of the nonmitochondrial contaminants removed, it was felt desirable to characterize and compare this preparation with other mitochondrial preparations. Table II lists the cytochrome con- tent of a crude mitochondrial preparation, the purified prepara- tion as described here, and a synaptosomal preparation (12), as estimated by two different techniques. It may be seen from the low temperature spectra of a fully reduced sample versus a fully oxidized sample (Figs. 2 and 3) that both the mitochondrial and synaptosomal preparation contained the normal complement of mammalian cytochromes absorbing at wave lengths which are similar to those of cytochromes in other tissues. The alpha re- gion of the spectra revealed a reduced cytochrome a and a3 ab- sorption with a maximum at 601 nm for the mitochondria and 599 nm for the synaptosomes, a reduced cytochrome b absorption apparent as a shoulder at 563 nm in both preparations and a re- duced cytochrome c peak with a sharp absorption maximum at 548 nm in the mitochondria and 549 nm in the synaptosomes. Cytochrome cl, which can only be visualized clearly in low tem- perature spectra appeared as a slight shoulder at 556 nm in the

by guest on Novem

ber 17, 2020http://w

ww

.jbc.org/D

ownloaded from

4726 Preparation of Rat Brain Mitochondria Vol. 245, No. 18

TABLE I Distribution of succinate and lactate dehydrogenase activity in subcellular fractions of rat brain

Fraction A was the nuclear and cell debris pellet formed on cen- dient procedure and gave rise to the supernatant (E) and the trifuging the homogenate for 2,000 X g, 3 min and Fraction B mitochondrial pellet (F). Fraction G was the Pellet F after a the corresponding supernatant. Fraction B was centrifuged for single washing with isolation medium. Succinate and lactate de- 12,500 X g, 8 min giving a supernatant (C) and a pellet (D). hydrogenase activity and protein were estimated as described Fraction D was then resuspended and subjected to the Ficoll gra- under “Methods.”

Fraction

Homogenate. .................... Low speed pellet (A). ............ Low speed supernatant (B). ...... B supernatant (C) ............... B pellet (D) ..................... Ficoll supernatant (E) ........... Ficoll pellet (F). ................. Washed mitochondria (G) ........

Protein

- 1: ;uccinate dehydrogenase activity

Total Total

mg % units”

784 100 2314 428 54.5 844 357 45.5 1470 189 24.1 610 141 18.0 602 91 12.0 392 20 2.6 82 10.2 1.3 71

-~ %

100 36.5 63.5 26.3 26.0 16.9

3.5 3.1

Lactate dehydrogenase activity

Total

units”

610 221 389 278 107

83.5 2.4 1.72

-

-

% 100 36.2 63.8 45.5 17.5 13.7

0.39 0.28

a One unit of succinate dehydrogenase activity = 1 nmole of formazan formed per mg of protein per min. b One unit of la&ate dehydrogenase activity = 1 pmole of NADH oxidized per mg of protein per min.

FIG. 1. Electron micrograph of a brain mitochondrial prepara- tion (Fraction G, see Table I) (magnification X 37,500).

sgnaptosomal preparation, whereas in the mitochondria it ap- peared as a split, but distinct, shoulder at 556 and 554 nm. The Soret region of the spectra showed the typical intense absorption of reduced cytochrome a and a3 at 443 nm in both preparations together with the cytochrome b and c shoulders at 430 and 415 nm which were only visible in the mitochondrial preparation.

The two methods used for estimating the total content of cyto- chromes in the mitochondria were the low temperature spectra using an intensification factor derived from control studies on crystalline cytochrome c and dual wave length spectroscopy using wave length pairs derived from spectra obtained from a split

- 1 Succinate to

actate dehydro- genase ratio

3.79 3.82 3.78 2.19 5.64 4.69

34.04 41.1

beam spectrophotometer at room temperature. It can be seen from Table II that there was good correspondence between these two methods. The cytochrome content per mg of protein in- creased 4- to 5-fold on purifying the mitochondria on the Ficoll gradient. There was also a change in the ratios of the cyto- chromes; the crude mitochondria contained relatively more cytochrome c than did the pure mitochondria. The levels and ratios of cytochromes were generally similar to those found in the mitochondria of other tissues, e.g. liver (20). The ratio of cyto- chromes in the synaptosomes was comparable to that in the purified mitochondria but the actual amount was less on a per mg of protein basis.

Additional experiments were carried out to estimate the total nicotinamide and adenine nucleotide content of these mitochon- dria. There was approximately 5 times as much NAD as NADP and a total adenine nucleotide content twice that of nicotinamide nucleotides (Table III). These data confirm the observations of Klingenberg, Slenczka, and Ritt (5) using a crude mitochondrial preparation, in which the total nucleotide content was consider- ably less than that found in mitochondria of other rat tissues (liver, heart, and kidney).

Previous workers (5) had indicated some difficulty in observing NAD(P) reduction on the addition of substrate (e.g. pyruvate) to their brain mitochondrial preparations. This function of these mitochondria was therefore investigated more closely by direct fluorometric read out and by extraction and enzymatic estima- tion of the nicotinamide nucleotides. Fig. 4 shows the degree of reduction of NAD(P) in various metabolic states of brain mito- chondria oxidizing pyruvate and malate; samples lvere removed at various points and enzymatically assayed for the nicotinamide nucleotides in order to provide confirmatory evidence of the fluorescence changes. Initially both NAD and NADP were greater than 90% oxidized. On addition of 2 mM malate, there was an increase in fluorescence which was not changed on adding 0.2 mM ADP. However, assay at this stage revealed that all the increase in fluorescence must be attributable to NADP reduction (83 % reduced) since there was no change in the NAD oxidation- reduction status. When 1 mM pyruvate was added there was a sharp increase in fluorescence followed by a slower increase which

by guest on Novem

ber 17, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of September 25, 1970 J. B. Clark and W. J. Nicldas 4727

TABLE II

Cytochrome content of rat brain mitochondria and synaptosomes

The wave length pairs and extinction coefficients used to calcu- late cytochrome concentrations from dual wave length measure- ments were: cytochrome 6, z$$~~“‘“““’ = 22; cytochrome c,

550 to 540 nm LAX = 19; cytochrome a, Z~~~,“~““““” = 24; and cyto-

445 to 450 nm chrome aa, &edox = 164 (17). The medium used was the 5 mM K+ buffer described under “Methods” and the protein con- centration was approximately 1.3 mg per ml. The cytochromes were oxidized by adding 10 PM rotenone and then reduced by add- ing a small quantity of dithionite. Cytochrome b was also re- duced and estimated by adding 10 rg of antimycin per ml and 5 mM succinate. The low temperature spectra were run in the same medium at protein concentration of approximately 5 mg per ml. In all cases the reference (oxidized) sample contained 10 PM rote- none and the measure (reduced) sample 10 .UM rotenone plus dithionite. Cytochrome content was calculated assuming an intensification factor equal to that of cytochrome c (la), and the extinction coefficients used were the same as above. Cytochrome c1 (a1 = 556 nm) was assumed to have parameters equal to those of cytochrome c (17). Each value is average of at least three separate estimations.

Preparation and method of estimation

Crude mitochondria Dual wave length spectro-

photometer. Ratio.

Low temperature spectra. Ratio.......................

Purified mitochondria Dual wave length spectropho

tometer.. Ratio.......................

Low temperature spectra.. Ratio......................

Synaptosomes Low temperature spectra. Ratio.

Cytochrome content

0.025 0.86

0.029 1

0.14 0.88

0.11 0.16 0.61 0.9

0.033 0.05! 0.59 1.05

leveled off at a new steady state (State 4). There was a small

0.076 0.029 0.027 2.62 1 0.93

0.29 0.17 0.11 1.7 1 0.61

0.12 0.056 0.033 2.1 1 0.59

increase in NADP reduction but most of the fluorescence was

the result of a 5-fold increase in N,4D reduction. On the ad- dition of 1 rnr,f ADP (State 3) there was a decrease in fluores- cence attributable to a decrease in the reduction of NAD+ to one-third of its State 4 value and a slight oxidation of NADP. When rotenone was added all the NADP was reduced but only 70% of the NAD. From these results, therefore, it was con- cluded that these mitochondria function in a qualitatively similar way to heart mitochondria (21).

Respiration Rates-Table IV shows the respiratory activities of these brain mitochondria in States 3 and 4 in media containing either 5 or 150 mM K+. The respiration rates were assessed in both these media since it has been known for some time that K+ has a stimulatory effect on the respiration of brain preparations (22) and, more recently, on that of mitochondrial preparations (23). As an indication of their integrity, these mitochondria routinely showed respiratory control quotients of 7 to 10 with pyruvate and malate and 3 to 4 with succinate. Assessment of the actual rates of oxygen uptake with each substrate was com-

443 548 1

Ao.D.=O.O4

FIG. 2. Liquid nitrogen temperature (77°K) spectra of the cytochrome content of purified mitochondria. The mitochondria were suspended in oxygenated 225 mM mannitol-75 mM sucrose-5 mM K+ medium described under “Methods” at a protein con- centration of 5 mg per ml. The reference cell was treated with 10 PM rotenone before freezing. The measure cell was the same as the reference cell except that a small quantity of dithionite was added before freezing.

_t

AA=O.Ol

;.’ +40nmk-

FIG. 3. Liquid nitrogen temperature (77’K) spectra of the cytochrome content of sgnaptosomes. The synaptosomes were suspended in the 225 mM mannitol-75 sucrose-5 mM K+ medium at a protein concentration of 5 mg per ml. All other conditions as in Fig. 2.

plicated by the presence of a small endogenous respiration present when malate and ADP were added.

Although pyruvate alone was not oxidized appreciably, in the presence of malate it was .3xidized rapidly in both State 3 and the uncoupled state with carbonylcyanide-p-trifluoromethoxy- phenylhydrazone. This is appropriate since pyruvate formed via glycolysis is thought to provide the main substrate for energy metabolism in brain (24). Succinate was oxidized at rates com- parable to pyruvate, and was inhibited by malonate, an inhibi- tion which could be 50% reversed by the addition of concentra- tions of malate equimolar with malonate. Citrate, isocitrate, oc-ketoglutarate, and glutamate were all oxidized in the presence of malate at rates about 50% that of pyruvate-malate or succi- nate. The oxidation of citrate is of interest since it has been re- ported that citrate permease activity is low in brain mitochondria (25). When the malate-dependent transport, of citrate was in-

by guest on Novem

ber 17, 2020http://w

ww

.jbc.org/D

ownloaded from

4728 Preparation of Rat Brain Mitochondria Vol. 245, No. 18

TABLE III

Nicotinamide and adenine nucleotide content of rat brain mitochondria

The nicotinamide nucleotide levels are the average values f S.E.M. for the experiment shown in Fig. 4. The adenine nucleo- tides were also measured enzymatically in the same 5 mM K+ buffer with the mitochondria respiring in State 4.

Total nucleotide

z(NAD + NADH) 2.48 f 0.07 z(NADP + NADPH). 0.52 f 0.03 z(ATP + ADP + AMP) 5.92 f 0.1

I IpM Rotenone ImM pP 1

r NAD+

NADH

NADP+

NADPH

NAD

NADP

A 0 C D E

n moles /mg Protein

2.49 2.42 1.86 2.21 0.71

0.11 0.12 0.68 0.25 1.52

0.41 0.10 a05 0.07 co.05

0.04 0.50 0.46 0.42 0.59

% Reduction

4.6 4.7 26.8 10.2 68.1

8.9 83 z-90 86 >92

FIG. 4. Changes in pyridine reduction associated with State 4 --) State 3 transition in rat brain mitochondria incubated with 1 mM pyruvate-2 mM malate. The medium used was the 5 rnM Kf buffer described under “Methods.” Mitochondrial protein (10 mg) was suspended in an initial volume of 3.0 ml and 0.5.ml sam- ples were taken for analysis of pyridine nucleotides at the indi- cated places (14). Pyr, pyruvate; Mal, malate.

hibited by 5 mM butylmalonate (26), little oxygen uptake was

observed. However, on addition of 5 mM malate, citrate oxida-

tion was partially restored (50%). Although acetylcarnitine was oxidized at rates comparable to isocitrate or glutamate, ace- tate itself was not oxidized significantly even in the presence of malate and carnitine. Likewise NADH, octanoate, and P-hy- droxybutyrate did not stimulate the endogenous rate of oxygen uptake even in the presence of malate. Monofluoroxalacetate, a malate dehydrogenase inhibitor (27), added in the presence of malate, inhibited the oxidation of endogenous substrate in these mitochondria but pyruvate, added subsequently, was oxidized at a rate which was 40% of the normal pyruvate plus malate rate.

TABLE IV

Effect of Kf on oxidation of substrates by rat brain mitochondria

The media used contained either 5 or 150 mM K+, and oxygen uptake followed polarographically as described under “Methods.” Approximately 1 mg of mitochondria was suspended in a final volume of 1 ml in each case. State 3 conditions were initiated by addition of 0.5 mM ADP. In each case, the endogenous State 3 rate was measured prior to addition of substrate.

Added substrate

None.................................. 0 0 0 3.6 + 2.5 mM malate.. 0 21 0 25

1 mh4 pyruvate. 0 10 0 14.5 + 2.5 mM malate. _. 17 106 23 170 + 2.5 mM malate + 10 PM FCCPa.. 94 145

2.5 mM malate + 1 mM FIOAA. 5 + 1 mM pyruvate. 39 39 66

2.5 mM malate + 10 mM acetylcarnitine 29 59 27 76 2.5 mM malate + 5 mM citrate.. 17 66 48 92 0.25 mM malate + 5 mM butylmalonate 0 0

+ 5 mM citrate.. 7 5 + 5 mM malate.. 20 56 + 1 mM pyruvate. 33 83 56 175

2.5 mM malate + 4 mrvr isocitrate.. 11 40 29 86 2.5 mM malate + 5 mM ol-ketoglutarate. 16 46 30 85 2.5 mM malate + 2 mM glutamate.. _. 11 65 26 90 2.5 mM malate + 10 mM acetate. 0 17 0 26

+ 1 mM carnitine.. 0 17 0 26 10 mM succinate.. 34 122 82 150

+ 5 mM malonate. 20 0 + 5 mM malate.. 66 83

2.5 mM malate + 1 mM octanoate, 0 0 20 2.5 mM malate + 1 mM P-hydroxybutyr-

ate................................ 0 0 37 0.5 mM NADH. 8 0

+ 2.5 mM malate.. 18 18 15 21 10 mM D,L-a-glycerophosphate. 56 66 46 46

+ 10 PM rotenone.. 48 25 10 pM rotenone. 0 0 0 0

f 10 m&r D, n-ol-glycerophosphate. 52 52 24 24

Rf Concentration in medium

Sm&xK+

3tate state 4 3

15011~ K+

a The abbreviations used are: FCCP, carbonylcyanide-p- trifluoromethoxyphenylhydrazone; FiOAA, monofluoroxaloace- tate.

This confirms the findings of Kun and Volfin (27). a-Glycero- phosphate was also oxidized by these mitochondria (28)

With the exception of cr-glycerophosphate, the respiratory rates in the 150 mM K+ medium were greater than those found in the 5 mM K+ buffer with all the above substrates. With either pyruvate-malate (Fig. 5) or succinate (Fig. 6) as substrates, there was a half-maximal stimulation of oxygen uptake in State 3 at approximately 20 mM K+. Ozawa et al. (23) previously re- ported a half-maximal stimulation of glutamate and succinate oxidation with 40 mM K+ in their brain mitochondrial prepara- tion. The rates in State 4 with both substrates also doubled on increasing the K+ concentration but more gradually. The addi- tion of valinomycin, an antibiotic which stimulates the K+-ac- cumulating mechanism of mitochondria (29), caused a stimula- tion of the State 4 respiration to a rate approaching that of State

by guest on Novem

ber 17, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of September 25, 1970 J. B. Clark and W. J. Niciilas 4729

5 120-

m st.3

w60 f

w _.

2 : .- 80- Vd 40 5 2

I

.; y-- 2

40.

-‘;--_

$ A st. 4

20 F

;I d

6 5 5 e E 0 0

0 5 IO 20 50 150

K+d

FIG. 5. Effect of K+ on oxygen uptake of rat brain mitochondria incubated with 1 mM pyruvate-2.5 mM malate in presence and absence of ADP or 5 ng per ml of valinomycin (Val). The media contained varying amounts of KCl; sucrose and mannitol were correspondingly adjusted to maintain isotonicity with the media described under “Methods.” Other conditions are the same as those in Table IV. St, state.

If .

. St 3

) 5 20 50 j+GO

K*mM

FIG. 6. Effect of K+ on oxygen uptake of rat brain mitochondria incubated with 5 mM succinate in the presence and absence of ADP. The endogenous oxygen uptake of these mitochondria was inhibited by addition of 10 PM rotenone, otherwise, the experi- mental conditions are those in Fig. 5. St, state.

150 . 140 . 6 150mM K’

5mM KC

Val nghl

FIG. 7. Effect of valinomycin on the oxygen uptake of rat

brain mitochondria incubated with 1 mM nvruvate DIUS 5 mM malate in media containing either 5 or 150 mM KC. Conditions are those described in Table IV. VuZ, valinomycin.

3 at that particular concentration of K+ (Fig. 5). The concen- tration of valinomycin used to achieve maximal stimulation was the same (4 to 5 ng per mg of protein) for both the 150 and 5 mM K+ media (Fig. 7).

To examine the hypothesis that K+ exerts a direct effect on the cytochrome system in brain mitochondria (23), low tempera- ture cytochrnme spectra were made in the presence and absence of

TABLE v

Effect of K+ on cytochronze oxidation in brain mitochondria incubated with 1 rnM pyruvate-5 mM malate

Steady state levels of cytochromes were calculated from 77°K spectra obtained with the trapped steady state technique (16). The pertinent wave lengths and extinction coefficients used are those in Table II.

Concentration of K+ in medium

Cytochrome .SmxK+ 150 rn~ K+

state 4 state 3 state 4 state 3

y* reduced

2 2 2 3 6 0 5 <5

21 0 4 0 22 0 <5 0 27 16 44 27

ADP with pyruvate-malate as substrate (Table V). In media containing either 5 or 150 mM K+, the addition of ADP caused the oxidation of the cytochromes. In the case of cytochrome b, ap- proximately half that reduced in State 4 became reoxidized in State 3. However, in both States 3 and 4 twice the amount of cytochrome b was reduced in the 150 mM K+ medium as in the 5 mM K+ medium (Table V).

DISCUSSION

It has been known for some time that brain mitochondria can be prepared relatively free from contaminating subcellular par- ticles by density gradient separation procedures (7, 30, and 31). However, in these teshtiiques the mitochondria are centrifuged at very high speeds in extremely hypertonic sucrose concentrations for long periods of time. Thus, the metabolic activities of the preparation are limited and not easily correlated with mitochon- dria from other tissues which are not subjected to such rigorous preparative procedures. Because of these drawbacks, many workers have preferred to prepare brain mitochondria with var- ious adaptions of the Schneider and Hogeboom (32) procedure for liver mitochondria (e.g. 2, 3, 5, 6, 33, and 34). This crude brain mitochondrial fraction has been shown to be grossly con- taminated with other subcellular particles and fragments such as myelin (8) and synaptosomes (4, 30). Indeed, the latter authors use this crude fraction as the starting point in their sep- aration of synaptosomes. The resulting contamination of the crude fraction by nonmitochondrial protein and enzymes could well explain the inconsistencies in the metabolic properties at- tributed to brain mitochondria in some of the above reports, e.g. the data of Klingenberg et al. (5) indicated that NAD was not reduced by brain mitochondria oxidizing pyruvate and malate. This observation can be explained by the presence of lactate de- hydrogenase in synaptosomal inclusions in the preparation. Other reports of lactate utilization by brain mitochondria (35) can be similarly explained. Once this contamination is removed, as in the present work (Table I), NAD is rapidly reduced by pyruvate and malate (Fig. 4).

Thus, to study the metabolism of isolated brain mitochondria, it is necessary to use a procedure which avoids the rigors of the sucrose density gradient separation method (hypertonicity and time) but succeeds in purifying the mitochondrial fraction. This paper reports a simple, rapid, discontinuous gradient method

by guest on Novem

ber 17, 2020http://w

ww

.jbc.org/D

ownloaded from

4730 Preparation of Rat Brain Mitochonclria Vol. 245, No. 18

using Ficoll for the purification of a crude mitochondrial fraction which provides brain mitochondria suitable for metabolic inves- tigation. The method described here differs from other pub. lished methods utilizing Ficoll in several important respects. Kurokawa et al. (10) developed a rapid method for isolating nerve ending particles with a gradient consisting of 0.32 M su-

crose, 3% Ficoll, and 13% Ficoll. Although this procedure has no known deleterious effects on synaptosomal metabolism, Stahl et al. (8) have shown that Ficoll solutions in excess of 8% yield mitochondria with poor respiratory activity. Stahl et al. (8) did not use a discontinuous Ficoll gradient but suspended the crude mitochondrial pellet in an 8% Ficoll medium and then cen- trifuged. The resulting mitochondria have an absolute require- ment for bovine serum albumin to show respiratory activity and control and, in addition, are able to oxidize exogenous NADH and cytochrome c. The mitochondria described here did not have these limitations.

Electron microscopy (Fig. 1) showed little contamination of this preparation by synaptosomes; the lactate dehydrogenase distribution (Table I) indicated that this synaptosomal contami- nant was less than 2%. Salganicoff and Koeppe (36) have re- ported that 75 y0 of the succinate dehydrogenase activity in brain is associated with nerve ending mitochondria. The results in Table I, therefore, are consistent with there being little contam- ination of this preparation by nerve ending particles or mito- chondria derived from them; this suggests the mitochondria described here are primarily of glial origin. Further inspection of the electron micrographs reveals only slight contamination with mitochondrial or other membranous fragments such as myelin.

Since this brain mitochondrial preparation is different from those described previously, the mitochondria were characterized with respect to cytochrome content, nucleotide content, and sub- strate utilization prior to metabolic investigations. The cyto- chromes found in these mitochondria were similar in their spec- tral properties, in total content on a per mg of protein basis and in their relative concentrations (Table II) as those found in other mammalian mitochondria (20, 37). The results are consistent with the data found on brain mitochondria by Sacktor and Packer (28) and Klingenberg et al. (5) if allowance is made for the rela- tive protein contents of the preparation but are inconsistent with the values reported by Williams (38) who used a different method of extraction and est,imation. This is further emphasized by the observation that, apart from cytochrome c, the cytochrome con- tents of the crude and purified mitochondria and synaptosomes have similar relationships to each other and only differ when they are related to the protein content of the preparation. The addi- tional cytochrome c in the crude mitochondrial fraction is prob- ably of extramitochondrial origin.

The total content of pyridine nucleotide was 2- to 3-fold greater on a per mg of protein basis (Table III) than those pub- lished by others (5, 28), and the ratio of NADP:NAD in these mitochondria was 30% greater than that reported by Klingen- berg et al. (5). This latter difference may be caused by the greater purity of the preparation or because of more efficient techniques of extraction and estimation.

The brain mitochondria consistently showed a small, but sig- nificant, oxygen uptake on addition of malate and ADP (Table IV). This oxidation of malate was accompanied by the forma- tion of citrate (1 to 2 nmoles min-1 mg of protein-‘) indicating that acetyl-CoA from endogenous sources was also being uti- lized. The presence of this malate-stimulated endogenous res-

piration accentuates the importance of the order of addition of substrates and ADP to the mitochondrial suspension. Indeed, one could observe an apparent oxidation of an added substrate when, in fact, this was the result of the endogenous respiration. Since in much of the earlier literature, there was no recognition of this fact, the resulting data concerning respiration rates are often uninterpretable (e.g. 3,23,28). It has been suggested that gluta- mate is the chief endogenous substrate for isolated brain mito- chondria (39,40). However, the mitochondria used in the latter studies correspond to the crude mitochondrial fraction described in this paper, and conclusions were based on measured amino acid contents of the preparation. We have observed that, while there is a significant amount of glutamate in this crude fraction, the purified mitochondria contain less than 5 nmoles per mg of protein.2 This is hardly sufficient to maintain the observed endogenous rates. It is more likely that fatty acids are the endogenous fuels utilized by isolated brain mitochondria.

The highest respiration rate and respiratory control of all the substrates tested was shown by pyruvate plus malate (Table IV). This is entirely consistent with current models of brain metab- olism in which 90% of respiration is accounted for by the prod- ucts of glycolysis (24). The absolute requirement of malate for pyruvate oxidation could have several explanations. The iso- lated mitochondria may be extremely depleted of oxalacetate, and malate serves as a source of this compound. In view of the observed enhancement of respiration by K+, perhaps the malate acts to facilitate the transport of K+ into the mitochondria. Harris, Cockrell, and Pressman (29) have suggested that Kf may penetrate the mitochondria as salts of substrate anions such as malate.

These mitochondria utilized a complement of substrates nor- mal for mammalian mitochondria. It is of interest to note that citrate and isocitrate were oxidized at rates comparable to that of cr-ketoglutarate or glutamate. These observations are incon- sistent with the reports by Chappell (25) that the tricarboxylic acids are not utilized very well by brain mitochondria. The rela- tive intactness of the mitochondria was indicated by the fact that, exogenous NADH was not oxidized.

The oxidation of 5 mM L-oc-glycerophosphate, which was ro- tenone insensitive, was unique among the substrates tested in that the respiration was lower in the 150 mM Kf than in the 5 mM Kf at the same a-glycerophosphate concentration. Pre- liminary experiments indicate that this may be caused by a change in the K, for a-glycerophosphate. Since the oxidation of a-glycerophosphate by other mitochondria is known to be acti- vated by Ca*+ or Mg2f (41), the possibility exists that exter- nally added K+ displaces these cations from the mitochondria. Preliminary observations indicate that externally added Mg*+ did not stimulate this oxidation. Furthermore, although the brain mitochondria were capable of Ca2+ accumulation and subsequent extrusion under the appropriate conditions, there was no move- ment of endogenous CaZf caused by either K+ or uncoupler as measured by the Aequorin technique3 (42).

It has been suggested that the K+-stimulated increase in oxi- dation of substrates by brain mitochondria is caused by a direct stimulatory effect of K+ on the electron transport system (23). Measurements of the state of reduction of cytochromes (Table V) by trapped steady state technique (15) indicated, at least in the

2 J. B. Clark and W. J. Nicklas, unpublished observations. 3 J. B. Clark, W. J. Nicklas, and A. Azzi, unpublished observa-

tions.

by guest on Novem

ber 17, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of September 25, 1970 J. B. Clark and W. J. Nicklas 4731

presence of ADP, no forward crossovers between O2 and cyto- chrome b that would support such a conclusion. In State 4, there is a reverse crossover between cytochromes c, cl, and b

when K+ is increased. This could be caused by an inhibition at this site by K+, or K+ might activate at some site between oxy- gen and cytochromes c, and cl. However, in both States 3 and 4, there was a forward crossover between cgtochrome b and sub- strate oxidation. Thus, the principal effect of increasing K+ in the medium is to enhance the availability of substrate for oxida- tion (Figs. 5 and 6). Valinomycin, which enhances the transport of Kf into mitochondria (29), when titrated with brain mito- chondria respiring in State 4 in medium with 5 or 150 mM K+, yielded maximal rates identical with the rates found with mito- chondria in State 3 at that respective K+ concentration (Fig. 7). Therefore, the rate of transport of K+ into the mitochondria would appear to be intimately associated with the achievement of maximal respiratory rates. This stimulation may well be asso- ciated with the classic observation that K+ stimulates respira- tion and glycolpsis in other brain preparations (22), as well as altering the metabolism of amino acids in brain slices (43, 44). Pressman and Lardy (45) in their studies on the potassium re- quirements of liver mitochondria proposed that K+ sensitivity resides in reactions closely allied with phosphate esterification. The activity of the multienzyme pyruvate dehydrogenase com- plexes isolated from mitochondria of bovine kidney and heart and pig liver has been shown recently to be regulated by phos- phorylation and dephosphorylation reactions catalyzed by an ATP-dependent kinase and a phosphatase (46, 47). Thus, the K+-stimulated oxidation of pyruvate by brain mitochondria may be a reflection of the state of activation of the pyruvate de- hydrogenase complex. It is interesting to conjecture whether the analogous ac-ketoglutarate dehydrogenase is similarly regu- lated. The more detailed mechanism by which the stimulation occurs with brain mitochondria is the subject of a forthcoming publication.

This report has concerned itself with the preparation and some of the properties characterizing rat brain mitochondria. This preparation has been shown to be metabolically active and sub- stantially free from nonmitochondrial contamination. It is hoped that the use of this preparation will help solve some of the perplexities which too often have been associated with the study of the metabolism of the central nervous system.

Achnomledgments-We wish to thank Dr. J. R. Williamson for his constant encouragement and advice during the course of this work. We are further indebted to Professor Britton Chance for fruitful discussion and crit.icism and to Dr. D. F. Wilson for his help and advice on the cytochrome studies.

1. 2. 3.

4. 5.

6.

7.

8.

9.

REFERENCES

CLARIS J. B., Fed. Proc., 29, 471 (1970). JOBSIS, F. F., Biochim. Biophys. Acta, 74, 60 (1963). OZAWA, K., SETA, K., TAKEDA, H., ANDO, K., H-~NDA, H.,

AND ARAKI, C., J. Biochem. (Tokyo), 69, 501 (1966). GRAY, E. G., AND WHITTAKER, V. P., J. Anat., 96, 79 (1962). KLINGENBERG, M., SLENCZKA, W., AND RITT, E., Biochem. Z.,

332, 47 (1959). CUNNINGHAM, R., AND BRIDGERS, W. F., Biochem. Biophys.

Res. Commun., 38, 99 (1970). SALGANICOFF, L., AND DE ROBERTIS, E., J. Neurochem., 12,

287 (1965). STAHL, W. L., SMITH, J. C., NAPOLITANO, L. M., AND BASFORD,

R. E., J. Cell Biol., 19, 293 (1963). BASFORD, R. E., in R. W. ESTABROOK AND M. E. PULLMAN

(Editors), Methods in enzymology, Vol. 10, Academic Press, New York, 1967, p. 96.

10.

11. 12. 13.

14.

KUROKAWA, M., SAKAMOTO, T., AND KATO, M., Biochim. Bio- phys. A&, 94, 307 (1965)‘.

MOORE. C. L.. J. Neurochem.. 15. 883 (1968) BRADFORD, H: F., J. Neurochkm.; 16, 675 (i969). ESTABROOK, R. W., AND MAITR.4, P. K., Anal. Biochem., 3,

369 (1962). WILLIAMSON, J. R., AND CORICEY, B. E., in J. M. LOWENSTEIN

(Editor), Methods in enzymology, Vol. 13, Academic Press, New York, 1969, p. 434.

15. 16. 17.

18. 19.

20.

21.

22. 23.

24.

KEILIN, D., AND HARTREE, E. F., Nature, 164, 254 (1949). WILSON’, D.‘F., Arch. Biochem. Biophys., 121, 757 (1967). WILSON. D. F.. AND EPEL. D.. Arch. Biochem. Biovhus.. 126.

83 (19ij8). ’ , I -1 ,

PENNINGTON, R. J., Biochem. J., 80, 649 (1961). GORNALL, A. G., BARDAWILL, C. S., AND DAVID, M. M., J.

Biol. Chem., 1’77, 751 (1949). ESTABROOK, R. W., AND HOLOWINSKY, A., J. Biophys. Bio-

them. Cytol., 9, 19 (1961). LANOUE, K., NICKLAS, W. J., AND WILLIAMSON, J. R., J.

Biol. Chem., 246, 102 (1970). KINI, M. M., AND QUASTEL, J. H., Nature, 184, 252 (1959). OZAWA, K., SETA, K., ARAI~I, H., AND HANDA, M., J. Biochem.

(Tokyo), 61, 352 (1967). ELLIOTT, K. A. C., AND WOLFE, L. S., in K. A. C. ELLIOTT, I.

H. PAGE. AND J. H. QUASTEL (Editors), Neurochemistry. Ed. 2, Cdarles C Thomas, Springfield, Illinois, 1962, p. I&:

25. 26.

27.

28. 29.

30.

CHAPP&LL, J. B., Brit. Med. &11.~24, 150 (1968). ROBINSON. B. H.. AND CHBPPELL. J. B., Biochem. Biovhus.

Res. Commun., i8, 249 (1967). . ”

KUN, E., AND VOLFIN, P., Biochem. Biophys. Res. Cmmun., 23, 696 (1966).

SACKTOR, B., AND PACKER, L., J. Neurochem., 9, 371 (1962). HARRIS, E. J., COCKRELL, R., AND PRESSMAN, B. C., Biochern.

J., 99, 200 (1966). DE ROBERTIS, E., PELLEGRINO DE IRALDI, A., RODRIGUEZ

DE LORES ARNAIZ, G., AND SALGANICOFF, L., J. Neurochem., 9, 23 (1962).

31.

32.

33. 34.

35.

36.

37.

NEIDLE, A., VAN DEN BERG, C. J., AND GRYNBAUM, A., J. Neurdcherk., 16, 225 (1969).

SCHNEIDER, W. C., AND HOGEBOOM, G. H., J. Biol. Chem., 183, 123 (1950).

BRODY, T. M., AND BAIN, J. A., J. Biol. Chem., 196, 685 (1952). ABOOD. L. G.. GERARD. R. W.. BANKS. J.. AND TSHIRGI. R. E..

Amer. J. ~Piysiol., 168, 728 ‘(1952). ’ ’ SACKTOR, B., PACKER, L., AND ESTABROOK, R. W., Arch. Bio-

them. Biophys., 80, 68 (1959). SALGANICOFF, L.. AND KOEPPE, R. E., J. Biol. Chews.. 243,

3416 (1968): ’ KLINGENBERG, M., in E. KUHN (Editor), Symposium tiber

progressive Muskeldystrophie, Springer-Verlag, Berlin, 1966, p. 488.

38. WILLIAMS, J. N., JR., Biochim. Biophys. Acta, 162, 175 (1968). 39. BELLAMY. D.. Biochem. J.. 82, 218 (1962). 40. COHEN, g. P:, in M. M. COHEN AND R. s. SYNDER (Editors),

Morphological and biochemical correlates of neural activity; Harrier and Row. New York. 1964, v. 212.

41.

42.

ESTABROOK, R. W:, AND SAC~TOR, ‘g., J. Biol. Chem., 233, 1014 (1958).

43.

44.

45.

46.

47.

CHANCE, B., AZZI, A., LEE, I. Y., LEE, C. P., AND MELA, L., in L. ERNSTER AND Z. DRAHOTA (Editors), Mitochondria struc- ture and function, Fifth Meeting of Federation European Biochemical Societu. Praaue. 1968. Academic Press. New York, 1969, p. 245.“’ ” ’ ’

BERL, S., NICKLAS, W. J., AND CLARKE, D. D., J. Neurochem., 16, 131 (1968).

BERL, S., NICKLAS, W. J., .~ND CLARKE, D. D., J. Neurochem., in press.

PRESSMAN, B. C., AND LARDY, H. A., Biochim. Biophys. Acta, 18, 482 (1955).

LINN, T. C., PETTIT, F. H., HUCHO, F., AND REED, L. J., Proc. Nat. Acad. Sci. U. S. A.. 64. 227 (1969).

WIELAND, O., AND SIESS, E., P&c. Nat. Acad.‘Sci. U. S. A., 66, 947 (1970).

by guest on Novem

ber 17, 2020http://w

ww

.jbc.org/D

ownloaded from

John B. Clark and William J. NicklasCHARACTERIZATION

The Metabolism of Rat Brain Mitochondria: PREPARATION AND

1970, 245:4724-4731.J. Biol. Chem. 

  http://www.jbc.org/content/245/18/4724Access 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/245/18/4724.full.html#ref-list-1

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

by guest on Novem

ber 17, 2020http://w

ww

.jbc.org/D

ownloaded from