THE JOURNAL OF Bmmorcnr. CHEMISTRY VoI.249, No. PO,Issue of October 25, PP. 6495-6504, 1974

Printed in U.S.A.

The Effect of the Hypocholesterolemic Drug Clofibrate on

Liver Mitochondrial Biogenesis

A ROLE FOR NEUTRAL MITOCHONDRIAL PROTEASES*

(Received for publication, December 17, 1973)

ADRIAN R. L. GEAR,$ ARLENE D. ALBERT, AND JANA M. BEDNAREK

From the Department of Biochemistry, The University of Virginia., Charlottesville, Virginia 22901

SUMMARY

The hypocholesterolemic drug clofibrate raises hepatic levels of mitochondria by over loo%, with maximum increase at 2 to 4 days after drug administration, and a slow rise con- tinuing at 10 days. This time response was similar to that seen after partial hepatectomy. However, several clear dif- ferences exist. During the doubling in mitochondrial con- tent, no change in particle size or specific content of DNA or RNA was observed, whereas during liver regeneration the former decreases by 50% and the latter increases by 300%. The half-life of 5.8 days for normal mitochondria was not altered significantly by clofibrate, using [guanidinio-WI- arginine. However, the initial rate of short term incorpora- tion of [35S]methionine into control mitochondria in viuo was 60% of that for drug-tested animals, while the in vitro effi- ciency of [WJleucine incorporation was reversed; mitochon- dria from drug-treated animals were about 61% as active as those from controls. Enzyme activity of drug-treated ani- mals was not significantly altered for inner membrane or ma- trix enzymes. However, a 30 to 40% fall in specific activity of three outer membrane enzymes was noted from 2 to 15 days. This resembles changes seen during early liver re- generation.

One striking action of clofibrate was to inhibit the neutral protease activity of mitochondria. Even at 2 days on clofi- brate, neutral protease activity was inhibited to 62% of the control value of 0.334 pmole of amino acid per hour per mg of protein. Inhibition was a maximum of 50% at 6 days, and was still at 63% at 20 days. Clofibrate added in vitro had negligible influence on neutral protease activity. _ The inhibition was independent of lysosomes since the protease activity of purified lysosomes was neither inhibited, nor was acid phosphatase activity in whole homogenates or the puri- fied mitochondrial fractions changed during drug treatment. The endogenous amino acid content of 0.032 pmole per mg

* This research was supported in most part by a subgrant from an institutional grant to&e University of Virginia by the United States Public Health Service (National Institutes of Health). , , Biomedical Sciences Support, BSS, 5-S05-RR07094-07, in minor part by the National Institutes of Health NHLl (HL-15289), as well as to A. D. A. from a research training grant of the United States Public Health Service (National Institutes of Health) GM- 01814.

1 To whom requests for reprints should be addressed.

protein remained constant during the stimulation in bio- genesis.

The results of this study on clofibrate support a hypothesis that net levels of hepatic mitochondria may be controlled by the activity of their own neutral proteases. A possible action is to degrade newly synthesized mitochondrial proteins, which may be structural or binding, such that a lowered protease activity would raise their levels and thereby increase mitochondrial content.

Since the introduction of the hypocholesterolemic drug clo- fibrate in 1962 (1, 2), an enormous volume of literature has de- scribed its useful clinical ability to lower serum levels of both cholesterol and triglycerides. Various hypotheses have been proposed for rationalizing this role. However, none are com- pletely satisfactory. For example, explanations of the mode of action of clofibrate range from decreased absorption of cholesterol and enhanced excretion (3, 4), through accelerated conversion of cholesterol to bile acids, inhibition of cholesterol synthesis (5), and a redistribution of cholesterol between plasma and tissue compartments (6, 7).

An early observation that hepatomegaly occurred during clo- fibrate administration (8) stimulated interest in liver function, particularly since this organ is so vitally concerned with lipid and cholesterol transport, as well as cholesterol oxidation to bile acids. Subsequent to this, a useful study was reported by Hess et al. (9), on some specific enzyme changes. They noted a dramatic g-fold increase in hepatic levels of glycerol phosphate dehydrogenase specific activity, a significant (74%) rise in cytochrome c oxidase, as well as a smaller 27% increase in catalase. Strangely, the latter microbody enzyme did not parallel urate oxidase, which fell by 58%. These observations correlated with an early elec- tron microscopy study by Paget (10) who showed an elevated number of normal-sized, but denser mitochondria. More lyso- somes were thought to exist, but this could have resulted from staining problems, since the later electron micrographs of Hess et al. (9) strongly supported an elevated number of microbody profiles, not lysosomes. Enzyme estimations (7) supported these latter conclusions. Interestingly, the lack of crystalloid in the microbodies correlated with the depressed urate oxidase activity seen after clofibrate treatment.

6495

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6496

In 1970 Kurup et al. (11) reported a detailed study on the specific ability of clofibrate to raise hepatic levels of mitochondria. They demonstrated that most of the increase in liver weight could be accounted for by mitochondrial protein alone, and that apart from the dramatic stimulation of glycerol phosphate dehydrog- enase activity (9), mitochondrial “quality” was not altered by the drug. A subsequent study with the related compound clo- fenapate (12), also demonstrated clearly that only hepatic levels of mitochondria were increased, not those of the nuclear, micro- somal, or supernatant fractions.

This striking ability of clofibrate to raise levels of hepatic mito- chondria provides a basis for a general hypothesis of the hypo- cholesterolemic and hyperlipidemic properties of the drug. It is known that although cholesterol conversion to bile acids occurs primarily in the cytoplasm, the final side chain cleavage to pro- pionyl-CoA and cholyl-CoA is a mitochondrial step (13, 14). Thus Kritchevsky et al. (15) described how clofibrate dramatically increased the ability of rat liver to oxidise [14C26]cholesterol to ‘4COz. However, when their data were normalized per mg of mitochondrial nitrogen, there was no difference between mito- chondrial fractions derived from control or clofibrate-treated rats. Consequently, the explanation for these observations probably lies in the ability of the drug to cause a net increase in the amount of mitochondriil protein per animal.

The above experiments might help explain the hypocholes- terolemic action of clofibrate by increase in mitochondrial choles- terol oxidation to bile acids and salts, and has been confirmed by others (16), but remains controversial (3). The former study in dogs was by analysis of bile obtained from a Thomas cammla, whereas the latter (3) involved analysis of fecal steroids and bile salts in humans. Another rationalization, still related to in- creased levels of hepatic mitochondria, could be reduced choles- terol synthesis. Several observations support this idea (5, 17) and specifically pointed to cholesterol synthesis from acetate and not mevalonate as being inhibited, but these findings have been questioned (for a discussion, see Ref. 5). Since the biosynthetic steps occur extramitochondrially (15), it could be asked how changes in the amount of mitochondria might influence choles- terol synthesis. Recently, however, Burch and Curran (18) found that the activity of mitochondrial acetoacetyl-CoA deacyl- ase was markedly increased in rats fed a clofibrate diet. The increase was about 65% on a total liver basis, and was quantita- tively significant in terms of the net cholesterol synthesis by liver. The authors pointed out that the consequence of the deacylase increase would be to reduce levels of cytoplasmic acetoacetyl-Cob necessary for mevalonate formation, and would fit in with the early observations of Gould et al. (5). They also presented evi- dence for a strong relationship between a reduction in digitonin- precipitable sterols and increased deacylase activity. The result could not be demonstrated in vitro, and thus depended on either a metabolite of clofibrate or some time-dependent induction of mitochondria. A recent report, however, describes separate pathways for ketone body formation and cholesterol synthesis (19).

A final involvement for the hypothesis of altered hepatic levels of mitochondria relates to the known hypotriglyceridemic action of clofibrate. One of the earliest changes in ensyme activity was that of mitochondrial glycerol a-phosphate dehydrogenase, which rose g-fold during drug administration (7, 9). A consequence of this rise could be to reduce cytoplasmic levels of glycerol 3-phos- phate needed for triglyceride and phospholipid synthesis. In- deed, Fallon et al. (20) reported a 46% lowering of hepatic levels for rats 2 weeks on clofibrate treatment, although these workers

thought the best hypolipemic explanation lay in an inhibition of the acyltransferase ensyme, and not with simple lack of necessary substrate.

This brief survey provides good evidence that the hypolipemic activity of clofibrate could well follow from a net increase in hepatic mitochondria. The effect of this could be either to ac- celerate cholesterol breakdown to bile acids, or inhibit its syn- thesis, or both, and to decrease glycerol 3-phosphate availability for triglyceride and phospholipid synthesis needed for very low density lipoprotein formation.

The research to be reported here was undertaken to investigate how clofibrate might act in altering levels of liver mitochondria, and to provide a firm basis for the hypothesis that the hypo- lipemic action of clofibrate is correlated with an increased liver content of mitochondria. The results of this study support the conclusion that clofibrate inhibits mitochondrial neutral protease activity, and thereby raises levels of newly synthesised protein for incorporation into mitochondria.

EXPERIMENTAL PROCEDURES

Chemicals

Cytochrome c (type III), DL-kynurenine sulfate, benzylamine, oxaloacetic acid, rotenone, p-nitrophenyl phosphate and Tris were all obtained from the Sigma Chemical Co., St. Louis, MO. NADH, NADPH, and ADP were purchased from P-L Biochemicals, Inc., Milwaukee, Wise. Glycylglycine; RNA, purified from Torula, DNA (from salmon sperm), and N-tris(hydroxymethyl)methyl-2- aminoethanesulfonic acid, were obtained from Calbiochem, Los Angeles, Calif. Lubrol-WX, a non-ionic detergent came from ICI Organics Inc., Provincetown, R. I. n-Glutamic acid came from Mann Research Laboratories, New York. n-Ascorbic acid, re- agent grade; orcinol, reagent grade; and diphenylamine, special indicator grade, were obtained from Fisher. All other reagents were analytical reagent grade, or the highest purity commercially available.

Isotopes

n-[U-‘4C]Leucine, n-[qua&&&o-lPC]arginine, and n-[%S]methi- onine were purchased from International Chemical and Nuclear Corporation, Irvine, Calif. Butyl-PBD ((4’~t-butylphenyl)-5-(4’- biphenyl)-1,3,4-oxdiazole), primary fluor, and scintillation tolu- ene and Biosolv (BBS-3) were products of Beckman Instruments, Inc., Fullerton, Calif.

Clofibrate

This was supplied in two forms which were the kind gift of Ayerst Laboratories, Inc., New York. All dietary studies were carried out using the ethyl ester; ethyl a-p-chlorophenoxyisobu- tyrate, which was incorporated into standard Purina rat chow at the content of 0.3’77c (w/w) by General Biochemicals, Chagrin Falls, Ohio. In vitro studies involved the sodium salt of clofibrate. which is water soluble.

Animals

Male. albino Sprague-Dawley rats were employed throughout the investigationandcame from Charles River Breeding Labora- tories. Wilminaton. Mass. The animals weighed from 150 to 200 g before beingslaced on the clofibrate diet. %ontrols were main- tained on normal rat chow until they were killed.

Apparatus

Mitochondrial counting and sizing was initially carried out with a Coulter Counter. model B. exactlv as described bv Gear and Bednarek (21). Final results were obtained using a-Celloscope LTH-112 (Particle Data, Inc., Elmhurst, Ill.) coupled to a labora- tory minicomputer and oscilloscope.1

1 A. R. L. Gear, E. J. Kuhnley, R. E. Bennett, and E. A. Craw- ford, submitted for publication.

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6497

Isolation of Mitoehondria dure (29) with 0.1,0.2, and 0.3 pmole of glutamate being employed as standards. Liver mitochondria were isolated by differential centrifugation

(22) from tissue homogenates in 0.25 M sucrose and stored at 0” at 50 mg of protein per ml. Heavy and light mitochondrial frac- tions were combined for the present studies. Preparation usually required about 134 hours from killing the rats, and all experiments with the mitochondria were usually completed within 4 to 6 hours.

Lysosomal Isolation

A lysosomal fraction was prepared according to the procedure of Sawant et al. (23), except that to increase the final yield and speed of isolation, the final washing, and centrifugation were omitted. Usually five to seven rats were killed to obtain a final fraction sufficient in amount and concentration for the subsequent assay of neutral protease activity.

Assay o.f Enzymic Activities

NAD+-Malate Dehydrogenase-The method described by Ochoa (24) was followed with Lubrol-WX being employed to re- lease the enzyme completely from the mitochondrial matrix.

Cytochrome c Oxidase-This was monitored spectrophoto- metrically at 25” by following the oxidation of reduced cytochrome c at 550 nm (22). Lubrol-WX was added at 0.2 mg per mg of mito- chondrial protein to activate fully the enzyme. Activity is ex- pressed as a first order reaction constant k-l.

Cytochrome c Reductases-NADH rotenone-sensitive, rotenone- insensitive, and succinate activities were assayed exactly as de- scribed by Sottocasa et al. (25). The rotenone concentration was 0.5 PM. Lubrol-WX was not employed since it inhibited the ac- tivities by about 20%.

Kynurenine Hydroxylase-This enzyme was assayed by the pro- cedure given by Hayaishi (26) with a Cary 14 spectrophotometer. The mitochondrial fractions were solubilized 15 min before assay by adding Lubrol-WX at a ratio of 0.3 mg per mg of mitochondrial protein, so as to minimize swelling effects.

Monoamine Oxidase-The method of Tabor et al. (27) was em- ployed in which benzaldehyde formation is monitored spectro- photometrically at 250 nm. Lubrol-WX was employed as above, to minimize lightscattering effects.

Acid Phosphatase-This activity was assayed by measuring the amount of p-nitrophenyl phosphate hydrolyzed at 37” (28). The specific activity of the enzyme is defined as micromoles of p- nitrophenol liberated per min per mg of protein, with an extinc- tion coefficient of 18.8 X lo6 M-l em+ being employed.

Neutral Protease Activity-The procedure which was developed is based on that described by Alberti and Bartley (29). An im- portant aspect of the assay is that concentrated mitochondrial suspensions must be used since proteolytic activity is not com- pletely linear with protein concentration (Table 2, Ref. 29). Thus, duplicates, where protein concentration was 10 and 26 mg per ml, were routinely employed. In this case, the q.,,, micromoles amino acid liberated per hr per mg of mitochondrial protein, is almost constant.

One result of Alberti and Bartley’s research was the observation that phosphate was a powerful inhibitor of proteolysis. Conse- quently, the influence of phosphate was tested in all assays. The following test tubes, in duplicate, were set up, 0.5 ml final volume.

Tube 0 123 4 5 6 TES* buffer, 4rnM ,

pH 7.4 KPO,, pH 7.4 0 0 0 2mbf 5mM 10mM 50mhi

Before incubation, 10 and 20 mg of mitochondrial protein were added to Tubes 0 and 1. resnectivelv. to serve as endoeenous blank. followed by 1 ml of 1.6 Nperchloric acid. Then 5 &d 10 mg of mitochondria were added to their respective tubes at the various phosphate concentrations, with incubation on a shaking water bath being for 1 hour at 37”. The reaction was stopped by addi- tion of 1 ml of 1.0 N perchloric acid. Following a 15-min period at O”, the tubes were centrifuged and the supernatants assayed for free amino acids, after neutralization and removal of the KClOd. The amino acids were estimated by a modified ninhydrin proce-

* The abbreviation used is: TES, N-tris(hydroxymethyl)- methyl-2aminoethanesulfonic acid.

After correction of the incubated tubes by the appropriate dilu- tion factors and endogenous amino acid blank, the results are expressed as a Qua, similar to those of Alberti and Bartley (29); that is, micromoles of amino acid liberated per hour per mg of mitochondrial protein.

The TES buffer selected for the assay of neutral protease ac- tivity was based on a survey of five “Goods” buffers (30). The results revealed that TES and N-2-hydroxyethylpiperazine-N’-2- ethanesulfonic acid were superior to the rest in maintaining incubation pH at the end of the hour, close to the original pH of 7.40. TES, however, gave the highest q,, of 0.277 pmole per hour per mg. Trial experiments also compared NaCl and sucrose with the known inhibition caused by phosphate. It was found that 50 mM PO, gave about the same Q~= as 250 rnM NaCl, and sucrose at 0.5 M only decreased the qaa by 50%, so the inhibitory effect is not simply a matter of tonicity.

All enzyme activities were measured at 25”. except where specially noted, with a Gilford automatic sampling- spectro- photometer coupled to a Honevwell Electronic 19. 10” recorder. The exception was kynurenine bydroxylase, which was monitored with a Cary 14 spectrophotometer. A Bausch and Lomb Spec- tronic 700, with automatic sampling cuvette, was used to obtain the extinction values for the ninhydrin amino acid assay.

Respiratory Control

This was measured by a polarographic technique with the suc- cinate medium described by Gear and Lehninger (31). It was routinely estimated as an index of mitochondrial structural in- tegrity immediately following completion of the mitochondrial preparation.

Estimation of Cytochromes (a + u3), b, cl, and c

These were measured using the 0.1 slide wire of a Cary 14 spec- trophotometer according to the technique of Williams (32), with more recently reported extinction coefficients for the appropriate wavelength pairs (33).

RNA Determination

This was done by the orcinol method of Schneider (34). The assay mixture contained orcinol dissolved immediately before use in concentrated HCl containing FeCl3. After heating for 20 min in a boiling water bath, the mixture was allowed to cool and then read against a blank at 660 nm. Standard curves were prepared from RNA purified from Torula corresponding to 10 to 160 pg of RNA.

DNA Estimation

The diphenylamine method according to Burton (35) was used. The assay mixture contained freshly dissolved diphenylamine in glacial acetic acid with subsequent addition of aqueous acetalde- hyde. After incubating for 16 to 20 hours at 30”, extinctions at 600 nm were read against a blank, with standard curve from 20 to 200 pg of salmon sperm DNA.

Isotope Incorporation Studies

Long Term in Vivo Estimation of Mitochondrial Half-life-Each rat was injected in the tail vein with 15 aCi of L-[guanidinio-W]- arginine in about 0.5 ml of isotonic saline. In one experiment only about 7 &i per animal was injected, due to a smaller purchase of the expensive isotope. Animals were then killed, always in pairs, from 2 days after isotope administration, up to 20 days, so as to obtain an accurate estimation of mitochondrial half-life with this non-reutilizable isotope (36).

Short Term in Vivo Incorporation-Here, since reutilization is not a problem, 10 &i of L-[+l]methionine were given intra- venously to each rat. Injections were accurately timed so that animals could be killed at 10, 20, 30, and 60 min after injection. This time scale is based on the experience of Beattie et al. (37) who demonstrated very rapid in vivo labeling of rat liver mitochondria.

In Vitro Amino Acid Incorporation-There has been considera- ble discussion as to the most suitable medium, especially concern- ing energy supply, for the in vitro amino acid incorporation into mitochondrial protein. A medium with an ATP-generating

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6498

system (38,39) was chosen, and incubation conditions were at 30” for up to 1 hour with most of the sterile precautions recommended by others (33). The amino acid mixture, minus leucine, was 2.5 mM with respect to alanine and had the molar ratios suggested by Roodyn et al. (40). Final mitochondrial concentration was 2.5 mg of protein per ml and L-[U-Wlleucine was added at 0.25 pCi per ml to initiate incorporation. The reaction was stopped at the various time intervals with trichloroacetic acid. The precipi- tates were washed twice with cold 10% trichloroacetic acid and prepared for scintillation counting as described below.

Scintillation Counting

The well washed precipitates from the various amino acid incorporation studies were dissolved in a small volume of 1 N

NaOH. The pH values of the resulting clear solutions were then adjusted to between 7.5 and 8.5 with 2 N acetic acid and the solu- tions diluted to give a final salt concentration of between 0.2 to 0.3 M. To 1 ml of these solutions were then added 10 ml of a scintillation mixture cont.aining (per liter) 8 g of butyl-PBD, and 15% of Biosolv (BBS-3). The solutions cleared immediately, but were not counted for several hours so as to allow for phosphores- cence decay. Counting was carried out in a Beckman LS-230 liquid scintillation counter. Results were expressed as counts per min per mg of protein actually contained in the final, neutral- ized solutions.

Protein Estimations

The ultraviolet absorption method of Murphy and Kies (41) was generally used on account of its speed and sensitivity. Some- times, however, the Folin method (42) was also employed. Bovine serum albumin was used as reference standard.

RFxiULTs

Injkence of Cl&n-ate on Liver Content of Mitochondria

Rats were given the drug at 0.3 ‘% (w/w) in their diet and mito- chondria isolated after 2, 4, 6, 10, and 15 days. The results illustrated in Fig. 1 demonstrate clearly the ability of clofibrate to induce a 100% increase in hepatic mitochondrial content within 10 days. Significance (p < 0.001) was high for all time intervals except, at 2 days, p < 0.1. The data were derived from nine separate experiments spread over 18 months with the same num- ber of control animals being sacrificed on each experimental day; the livers from two animals were always pooled before mito- chondrial isolation. Not shown are changes in liver weight. However, these have been extensively documented (8, lO-12,17), and in general for rats a 25 to 40% increase in weight occurs after 2 weeks on this diet. Also not shown are data obtained at 20

: c0rltrd

FIG. 1. The influence of clofibrate on the mitochondrial yield per animal. The data are derived from nine series of experiments, with two rats being killed at each experimental period and the number of control animals equaling the diet-treated ones.

and 30 days. A slow but steady increase relative to the control animals still occurred at these times.

The yields per animal are lower than those reported by Kurup et al. (II), but this may reflect different rat strains. One im- portant question to be asked was whether a change in the yield of isolated mitochondria could be responsible for some of the observed results. This was tested by assaying for total cyto- chrome c oxidase activity in the nuclear, mitochondrial, and supernatant fractions. Both diet and control animals gave es- sentially identical percentage yields in the three major fractions (Table I). This observation makes very unlikely that differen- tial mitochondrial fragility is responsible for the higher net yields obtained on clofibrate treatment. The total yield of cytochrome e oxidase per animal (Table I) also increased essentially in pro- portion to the recovered mitochondrial protein yield. A near doubling in the specific activity of homogenate cytochrome c oxidase was reported by Hess et al. in 1965 (9). In addition, Kurup et al. (11) revealed that the specific activity of succinate oxidase either per mg of liver protein, or per g wet weight of liver, increased dramatically and was about twice in the latter case. All of these results then confirm the specific increase in total liver mitochondria.

Mitochondrial Respiratory Control

This was tested on the same preparations used for Fig. 1, and the data are given in Table II. It may be noted that at 6 days on clofibrate there was a significant increase (p < 0.005) in respiratory control to a value of 5.52. This could well be related to a minimum in neutral protease activity, which will be presented later. The increase in respiratory control is reminiscent of Kurup et al. (ll), who did not present any statistical information, as well as in a more recent study of kidney cortex mitochon- dria (43).

Mat& and Inner and Outer Membrane Changes

It had been previously demonstrated (44) that during a period of massive mitochondrial biogenesis such as during liver regenera- tion, that a significant decrease in the specific activity of outer mitochondrial membrane enzymes occurred. On the other hand, mat,rix and inner membrane enzymes were constant. Interest- ingly, when rats were placed on the 0.3% (w/w) clofibrate diet,, a similar decrease in the outer membrane enzymes, monoamine

TABLE I Total cytochrome c oxidase activity in liver homogenates and

enzyme recovery in isolated mitochondrial fractions Results are expressed as percentages. The total activity per

animal represents the average from two separate series with clo- fibrate; one being direct measurement by summation of the total nuclear, mitochondrial, and supernatant activities; and the second from knowledge of cytochrome c oxidase specific activity (Table III), the total recovered mitochondrial protein (Fig. l), and the percentage of enzyme recovery in the mitochondrial fraction.

Days on clofibrate Control

2 4 6 10 15 --

Total homogenate activity (% of control). 100 108 174 196 202 216

Mitochondrial yield (To total homogenate enzyme ac- tivity) . 4G 51 40 49 48 58

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6499

TABLE II Respiratory control of mitochondria during clofibrate treatment

Eight separate series of experiments were carried out with two rats being killed at each time interval. Sodium succinate was the

substrate (31) and the respiratory control is defined as the ratio of the rate of respiration in the presence of phosphate acceptor (State 3) to that following, or in its absence (State 4).

AIlhal

Control Clofibrate-treated

Time (days)

2 4 6 10 1.5

3.20 f 1.39 3.66 f 0.42 3.40 f 0.70 3.82 f 1.00 3.48 f 0.97 3.96 f 1.42 3.82 f 0.52 5.52 f 1.42 3.90 f 0.51 4.39 f 0.88

p > 0.1 p> 0.1 p < 0.005 p > 0.1 p > 0.1

OL.‘,“““““‘l 0 5 10 15

Days on cloflbrate

FIG. 2. Outer, mitochondrial membrane enzymes during clo- fibrate treatment. Monoamine oxidase, (0); kynurenine hy- droxylase, ( l ) ; and rotenone-insensitive, NADH, cytochrome c reductase (A); were assayed as described under “Experimental Procedures.” The data are expressed as a percentage of the con- trol specific activity for each enzyme and represent the mean from two complete series of experiments.

TABLE III Cytochrome Content during Clofibrate Treatment Mitochcndrial inner membrane and matrix enzymes

during clo$brate treatment

The control specific activities are given for the four enzymes

and the results are expressed as percentages of these values.

The above results suggest that only outer membrane specific activities are changed during the large increase in liver mito- chondrial content, and that inner membrane and matrix enzymes remain constant. To co&m the latter findings, a direct analy- sis of the respiratory chain cytochromes was carried out. The results in Table IV show that no significant variations occurred in the specific content of the various cytochromes.

Enzyme Control specific activity

Cytochrome c oxidase 90 k-l Succinate cytochrome 0.15 Fmole/

c reductase min/mg Rotenone-sensitive 0.041 pmole/

NADH-cytochrome min/mg c reductase

NAD+ - malate de- 0.50 pmole/ hydrogenase min/mg

Days on clofibrate

2 -.

104

85

268

77

4 --

86 93

110

122

6 -.

90

79

122

10 15 20

85 82 82 65 105 90

95 68 81

89 92 94 94

oxidase, kynurenine hydroxyla.se, as Gel1 as rotenone-insensitive, NADH cytochrome c reductase was noted (Fig. 2). The final average reduction in specific activity for the three enzymes was by about 35%. As with the regenerating liver system, the specific activity of the typical matrix and inner membrane en- zymes did not alter significantly from the control values (Table III), except for the notable exception of rotenone-sensitive, NADH cytochrome c reductase at 2 days. Here the activity

OV~hdl

3.61 f 0.96

Mitochondrial cytochrome content during clofibrate treatment

The contents were estimated from the differences in extinction at the appropriate wavelength pairs, between oxidized and re-

duced mitochondrial suspensions, as described under “Experi- mental Procedures.” The results are the means from two com-

plete sets of experiments.

Control (0) 2

5 7 9

12

15 20

Specific cytochrome content

0.201 0.105 0.105 0.162 0.193 0.108 0.125 0.163

0.232 0.119 0.123 0.178 0.178 0.095 0.093 0.164 0.212 0.106 0.094 0.159

0.215 0.111 0.113 0.109

0.187 0.117 0.103 0.164

0.171 0.100 0.097 0.151

was 168 y0 above the control, comparable to the 147 ‘% increase noted at 2 days after partial hepatectomy (44). The significance of this is not clear.

Mitochondria DNA and RNA Content

It is known from early work with the regenerating rat liver system (45) and later reports (46-50), that the mitochondrial content of DNA and RNA increases several fold at 2 to 4 days after partial hepatectomy. This period preceded the large rise in liver mitochondrial content. In view of the similarities in time course, and net mitochondrial increase during liver regenera- tion (22, 45) and clofibrate treatment (Fig. l), it was anticipated that an increase in the specific content of mitochondrial nucleic acids might also take place. However, the data (Table V) reveal no significant changes in the mitochondrial specific content of DNA and RNA; the former remaining close to the control value of 0.304 pg per mg of protein, and the latter remaining near 7.56 pg of RNA per mg of mitochondrial protein. The whole problem of contamination by non-mitochondrial nucleic acids in the iso- lated fractions has been carefully considered (48) and, based on the use of nucleases, electron microscopy, washing procedures,

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6500

Mitochondrial content of DNA and RNA during clojibrate treatment

The nucleic acid content of well washed mitochondrial suspen- sions was estimated as described under “Experimental Proced- ures.” Results represent the means of two series of experiments.

All values are expressed as micrograms per mg of mitochondrial protein.

TABLE V TABLE VII Half-life of mitochondria from clofibrate-treated rats

Three series of experiments were carried out and the half-lives determined from a linear regression analysis of plots of log counts per min per mg of mitochondrial protein, against time. Animals were killed at 2- or 3-day intervals, beginning 2 days after ad- ministration of the isotope, which itself was begun 2 days after the clofibrate. The last isotope “day” was usually 18 or 20 days after diet commencement. Standard deviations are also given.

Days on clofibrate Nucleic acid

0 2 5 7 9 12 1.5 20 -~-~~~___~

DNA . . 0.304 0.290 0.336 0.283 0.242 0.399 0.301 0.325 RNA 7.56 8.74 8.33 5.67 7.38 7.67 7.62 5.13

l-~/n (days) y intercept Correlation coeficient

Control Clofibrate-treated

cpm/mK protein

5.8 f 1.5 43.8 f 5.6 -0.989 5.2 xt 0.9 41.0 f 3.0 -0.951

p > 0.1 p > 0.1 TABLE VI

Mean mitochondrial size during clofibrate treatment

Resistive particle sizing was done with a Coulter counter model B and subsequently with a Celloscope attached to a laboratory minicomputer as described under “Experimental Procedures.”

Days on diet Measurement

0 2 4 6 10 15 ---

Diameter, c.. . . . 0.98 0.92 0.97 0.94 1.02 1.04 Volume, pa. . 0.49 0.41 0.48 0.43 0.56 0.59

density gradient centrifugation, and marker enzymes, was shown to be of minor significance compared to the major changes during liver regeneration.

Mitochondrial Size

During liver regeneration it has been demonstrated that mito- ehondrial volume decreases by 50% 2 to 3 days after partial hepatectomy (48, 49), followed by a rapid return to the control volume. Using the resistive particle counting techniques de- veloped by Gear and Bednarek (21) as well as a newer one in- volving a laboratory minicomputer, mitochondrial suspensions were sized for up to 15 days on clofibrate. In contrast with the dramatic decrease in mitochondrial size during liver regeneration, no changes were evident (Table VI). Thus, the nucleic acid changes and mitochondrial size differentiate clearly the two ex- perimental situations where massive increase in liver content takes place.

0 10 20 30 40 50 60

Time after injection (mln)

FIG. 3. The short term incorporation of [%]methionine in viva. Two series of experiments were carried out, and the animals were sacrificed after 2,6, and 10 days on clofibrate. In one experi- ment, a 17.day pair was also killed. Specific activities as counts per min per mg of mitochondrial protein were averaged from each of the time intervals on diet. Finally, the specific activities of the control animals are expressed as a percentage of the clofibrate- derived mitochondria. The mean initial specific radioactivity (lo-min incorporation time) in absolute terms was 245 cpm per mg of mitochondrial protein for the drug-treated animals, and 168 cpm per mg for the controls.

Isotope Incorporation Studies

Long Term, Half-life Estimation-The use of the non-reutiliza- ble guanidinio group of arginine enables accurate estimates of mitochondrial half-life to be gained (36). Three complete series of experiments, including controls, were carried out with [guani- dinio-Wlarginine and the results are presented in Table VII. The difference between the control half-life of 5.8 days was not significantly different from that of the clofibrate-treated animals of 5.2 days. These values agree well with previous work (36) and also where reutilization was avoided with [r4C]carbonate (51) when the half-life for whole mitochondria varied between 4 and 6 days. They are quite distinct from workers who failed to use a non-reutilizable isotope, such as leuoine, (52), and obtained longer half-lives, near 8 to 10 days.

crease in the synthesis constant. Consequently, the rate of short term protein synthesis was analyzed. [a5S]Methionine was used for this study, and rats were sacrificed at 10, 20, 30, and 60 min after tail vein injection, since it is known that uptake is a very fast process (37,53). The present study with clofibrate is shown in Fig. 3, with the control data being expressed as a percentage of the specific radioactivity of the clofibrate mitochondria at the various time intervals. The intriguing result was obtained that clofibrate does indeed cause a stimulation of mitochondrial pro- tein synthesis, but that this was marked only at the shortest in- corporation times. The significance of this finding will be con- sidered later under “Discussion.”

Short Term in Vivo Incorporation-Since clofibrate did not alter the mitochondrial half-life, or decay constant, another way in which net change in mitochondrial levels could occur is via in-

Short Term in Vitro Incorporation-Many investigators have described the in vitro ability of mitochondria to incorporate radio- active amino acids into acid-precipitable material (3%40), but there is discussion as to exactly what is being labeled (54). In any event, the efficiency of in vitro labeling in mitochondria from clofibrate-treated animals was tested, especially since mitochon- dria from regenerating rat liver are more active at in vitro

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6501

1200

c ‘ZlOOO ‘j k

E” 000 \

E

2 - 600

s .- 74 b 400 La 8 .F 3 200

3 I

T 2-4 0 1

0 10 20 30 40 50 60

Incubation time (min)

FIG. 4. In vitro mitochondrial protein synthesis. Ten rats were placed on the clofibrate diet and then killed at 2,6,10, and 20 days. Two control animals were killed on each experimental day and in vitro protein synthesis with [W?]leucine was carried out as described under “Experimental Procedures.” Resulta are ex- pressed se counts per min of [Wlleucine per mg of mitochondrial protein, with the 5 experimental days being averaged together.

incorporation than controls (48, 55). The results of incorpora- tion at 10,20,30, and 60 min are given in Fig. 4, and show that, in sharp contrast to the in uiuo situation (Fig. 3), mitochondria from clofibrate-treated animals were on the average 61.2% as efficient as their controls. The data represent the average rates of incorporation at 2, 6, 10, 15, and 20 days on clofibrate. A noteworthy observation, before averaging the data, was that there was no difference between diet and control animals at 2 days. The inhibitory effect became apparent only after this time.

Znjhence of Clojibrde 4 Mitodwndrial Neutral Protease Activity

Mitochondrial neutral proteases have been described and their properties investigated (29, 56, 57), but, apart from speculation (58), no clear role has been ascribed to them. It was thought that they might be involved in mitochondrial turnover. To consider this possibility, their activity was studied during a period of massive change in liver content of mitochondrii, such as be- tween 2 and 10 days of clofibrate treatment.

When neutral protease activity was assayed in mitochondria from clofibrate-treated animals, it was tested at 2, 4, 6, 10, 15, and 20 days after initiating the diet. Four complete series of experiments were carried out, with controls, and two rats being killed at each time. Each test was run at 0,2, 5, 10, and 50 rnM phosphate to monitor the pattern of phosphate inhibition. The mean values from each of the six time intervals for increasing phosphate levels are shown in Fig. 5. The control mitochondria in the absence of phosphate is set at lOO$& The absolute value for this poo was 0.334 f 0.075 pmole of amino acid produced per hour per mg of protein at 37”. The absolute qoo for the clofibrate- derived mitochondrii was 0.206 f 0.059 rmole (p < 0.001). These data then demonstrate the powerful in viva inhibition that clofibrate has on the mitochondrial protease activit.y,

The time course during continuing clofibrate administration for protease activity is illustrated in Fig. 6. The data represent the mean percentage of inhibition from each of the phosphate con-

- 80 < 0

- t\

>, c > 60

clofibrate z 01’. ’ I 025 10 20 30 40 50

Phosphate ( m M 1

FIG. 5. The effect of clofibrate on mitochondrial protease activity at increasing phosphate concentrations. The data repre- sent the over-all means from four series of experiments with ani- mals being killed at 2,4,6, 10, 15, and 20 days of clofibrate treat- ment. It is given with the control in the absence of any phosphate being set at 100%. The absolute value in the absence of phosphate for the control was 0.334 f 0.075 pmole of amino acid produced per hour per mg of protein. In the presence of clofibrate, it was 0.206 f 0.059 mole.

8eoC\ 4 i60. w--- --I 0 2 4 6 a 10

Days on Clofibrate

12 14 16 20

FIG. 6. Mitochondrial neutral proteinase activity over a period of 20 days of clofibrate treatment. The same experimental format as given in Fig. 5 was used.

centrations. Once again, the striking inhibitory ability that clofibrate possesses at all times during drug treatment is evident, with a maximum 50% inhibition occurring at 6 days. It may be noteworthy that this was the one time the mitochondria had a highly significant increase in respiratory control (Table II).

The endogenous amino acid content of mitochondrii from all the preparations used for the control neutral protease measure- ments was 0.0324 f 0.009 pmole per mg of mitochondrial pro- tein, and 0.0319 f 0.008 mole per mg (p > 0.1) for the mito- chondria from clofibrate-treated animals.

Having obtained these results for an in viva action of clofibrate on protease activity, it was necessary to examine whether the drug might inhibit in vitro. Thus, the sodium salt of clofibrate, which is both water-soluble and the form the drug exists in plasma, was added directly to mitochondria at various levels. The results in Fig. 7 do show that there is a slight inhibition, but no more than expected from NaCl. Consequently, it is likely that the in utio inhibition is either the result of a biotransforma- tion of the drug or some secondary effect through another effector, either of low or high molecular weight.

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

There has been some discussion whether the neutral protease activity associated with mitochondria might simply reflect re- sidual lysosomal action at neutral pH due to the presence of contaminating lysosomes (56-58). However, the published evi- dence strongly supports the intrinsic mitochondrial origin of the neutral proteolytic activity. Nevertheless, to test whether lyso- somal artefacts might conceivably cause the earlier results (Fig. 6), acid phosphatase activity was measured in both the original homogenates and final mitochondriil suspensions. The results (Table VIII) reveal that there was neither a change in homog- enate specific activity, nor in the mitochondrial fractions, to parallel in any way the consistent and highly significant drop in neutral protease activity (Fig. 6).

A final set of experiments was carried out on the neutral pro- tease activity of a partially purified lysosomal fraction (23), to see whether the clofibrate-sensitive activity could be of mito- chondrial origin. The data in Table IX show that there is in- deed activity in the lysosomal fraction. However, it is less than that seen in control mitochondria of 0.334 I.cmole per hr per mg of protein, and this makes it unliiely that the higher mitochon- drial activity is the result of lysosomal contamination. Also it may be noted in Table IX that lysosomes from clofibrate-treated animals did not reveal a decreased neutral protease activity. If anything, activity was increased, although not significantly. Consequently, it is felt that these findings, plus the acid phos- phatase levels in the mitochondriil fractions mentioned above, give weight to the validity of the clofibrate-sensitive neutral pro- tease being truly mitochondriil.

100

- 80 < D

0 2 4 6 8 10 50

Salt (mM )

FIG. 7. The in vitro influence of clofibrate on neutral proteinase activity. The sodium salt of clofibrate was present in the incuba- tion media at, the various concentrations indicated, and the re- sults, expressed as percentages, are derived from three different mitochondrial preparations. Also shown, for comparison, is the inhibitory influence of phosphate.

DISCUSSION

The results of the research reported here support a hypothesis that mitochondrial levels may be determined by the activity of their own neutral proteases. The system studied, namely, that where the hypocholesterolemic drug clofibrate stimulates a large net increase in the amount of hepatic mitochondrii, revealed neutral protease activity to be consistently inhibited by about 40%. At the same time, although the mitochondrial half-life of 5 to 6 days was not influenced by the drug, the rates of short- term incorporation of radioactive amino acids in tie were stimu- lated by about 50%. These finclings led to the idea that a low- ered activity of neutral protease could increase the amount of the most rapidly synthesized, or newly incorporated protein, in mito- chondrii. Previous research (37, 53, 59), has shown membrane or structural proteins to have the highest specific radioactivity within 5 min of isotope injection. It was specifically at these early times that the most stimulation in protein synthesis was seen (Fig. 3). Consequently, it is proposed that the steady state levels of these structural or membrane-binding proteins determine the final steady state levels of total mitochondrial protein.

Evidence from earlier work (37,53,59) has indicated that mito- chondrial membrane proteins are preferentially labeled by iso- topes over the bulk or matrix proteins. What is then possible is that the availability of an intriiic bmclmg protein(s) could regulate whether additional proteins are incorporated into the membrane. In this context it is now known that three of the subunits of cytochrome c oxidase are synthesized by mitochon-

TABLE IX Neutral protease activity of purified lysosomal fractions

during clojibrate treatment The data is derived from 2,5, and 6 days of clofibrate treatment,

and compared against three sets of controls; five to seven rats were killed for each lysosomal purification, and the livers were pooled. Neutral protease activity in the absence of any phos- phate was assayed as described under “Experimental Proced- ures.” Acid phosphatase was also monitored in both the intial liver homogenate and final lysosomal fraction to obtain an indica- tion of the relative purification.

Fraction

Neutral protease: Qua (pmole of amino acid liberated/hr/mg protein). . . .

Purification factor: ratio of acid phosphataae specific activity in fraction compared to initial ho- mogenate...................... 6.8 f 2.2

-

--

/

-

Clofibrate lysosomes

0.27 i 0.12

5.0 + 1.8

TABLE VIII

Acid phosphatase activity during clofibrate treatment, measured in both homogenates and mitochondrial suspensions

Activity was measured by the hydrolysis of p-nitrophenyl phosphate as described under “Experimental Procedures.” All values are expressed as micromoles of substrates hydrolyzed per min per mg of protein.

I Tie on clofibrate (days)

Suspension

Mitochondria

Control 2 4 6 10 13

0.032 f 0.005 0.021 0.041 0.039 0.025 0.026

Homogenate 0.259 f 0.023 / 0.258 1 0.258 1 0.250 ( 0.264 1 0.281

Over-all mean

16 20

0.022

.255

0.025 0.028 f 0.008 p> 0.1

0.263 0.261 f 0.011 p > 0.1

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6503

drial ribosomes (60), and in their absence the remaining three cytoplasmic subunits are not incorporated. Similarly, compo- nents of mitochondrial ATPase which are synthesized on cyto- ribosomes will not be incorporated into the inner membrane un- less the appropriate binding protein is available (61). The present investigation of mitochondrial neutral protease activity provides an experimental rationaliiation for the concepts just discussed. Namely, the protease may have a higher affinity for or (because of its intramitochondrial localization) will preferen- tially degrade those proteins being most rapidly synthesized, which include the membrane-binding proteins considered above. The effect of this for clofibrate, which inhibits, although not in vitro (Fig. 7), the mitochondrial protease, would be to raise the steady state levels of the binding protein, or proteins. An ob- servation of Alberti and Bartley (56) is highly significant in this context. They found that when chloramphenicol inhibited in vitro mitochondrial protein synthesis, amino acid production de- creased. Thus, there is a direct relationship between rate of synthesis of new protein, which in vitro is only membrane-linked (40,54,62), and the rate of protein breakdown. In other words, the protease preferentially acts on newly synthesized membrane protein. It is possible to conjecture that some of these proteins may be proteolipids (63), as evidenced by electrophoretic be- havior and high specific radioactivity after in vitro labeling (54).

A few additional comments concerning the neutral protease activity are in order. It was mentioned under “Results,” but should be emphasized again; namely, there is a variety of evidence that lysosomes, either broken or whole, do not contribute to the neutral proteolytic activity (56, 57). This conclusion was based on differential solubilization, inhibitors, protective and swelling agents, electron microscopy and, finally, Sephadex chromatog- raphy of the solubilized proteases. Also the present research (Tables VIII and IX) gives additional weight to the assumption that clofibrate-sensitive protease is mitochondrial and not lyso- somal in origin. An earlier observation (29) that mitochondrial disruption increased the qaa, either when sucrose was present or when phosphate inhibition was tested, suggested that only a fraction of the total mitochondrial protein was accessible to di- gestion. Thus, the qao in uiuo is likely to be less than for isolated mitochondria; and, in addition, intracellular phosphate which is about 5 mM, plus other inhibitory anions, an osmotic strength of some 320 mosmolar (29), and substrate, will all combine to lower the qaa. If one assumes it to be similar to the inhibition seen at 50 mu phosphate (Fig. 5), then it is possible to calculate how long it would take the protease to digest completely 1 mg of mito- chondria. For control mitochondria a qsa of 0.062 gives a “life- span” of 6.7 days, and for clofibrate the qas of 0.043 yields 9.7 days.

These values are for a turnover faster than indicated by the normal mitochondrial half-life of 4 to 6 days (51). They prob- ably mean that mitochondrial turnover does not involve the neutral proteases, a concept supported by the results with clo- fibrate. However, this is not ruled out completely, and whether the proteases or lysosomes play the major role in mitochondrial turnover remains a fascinating problem.

Some recent investigations have been concerned with the in vitro effects of clofibrate on liver metabolism. For example, Panini (64) reported that the sodium salt at 1 rnM inhibited State 3 respiration by 50%, and a similar study (65) revealed essen- tially the same basic finding. Subsequently, a more complete report (66) documented two sites of interaction with the mito- chondrial respiratory chain. Finally, a very recent study sug- gests a highly specific inhibition action on certain mitochondrial

permeasesa Thus, clofibrate definitely possesses inhibitory prop- erties toward mitochondria and thus we might rationalize why the in vitro (Fig. 4) effect on protein synthesis was the reverse of that seen in uivo (Fig. 3). An alternative, and probably more likely possibility, is that isolated mitochondria lack cytoplasmic factors such as an exchange protein (67) needed for normal pro- tein synthesis. There is consequently a great danger of extrap- olating in vitro results to the in VCO situation.

One of the original stimuli for undertaking the research on clo- fibrate was to determine whether a system, where there was a large net increase in liver mitochondria, behaved similarly to that seen after partial hepatectomy (22,44,45). The time course and net mitochondrial increase are about the same in both situations, but several clear similarities and contrasts exist. Interestingly, the outer membrane enzymes decreased in specific activity in both cases (44, Fig. 2) ; although for regenerating liver the de- crease was only temporary. Another situation in which outer membrane activity declined was after thyroxine treatment of normal rats (68). It has been proposed that one of the modes of action of clofibrate could be to displace bound thyroxin. The effect is reasonable since, in both thyroxin- and clofibrate-treated rats, glycerol phosphate dehydrogenase activity increased dramatically (9, 68) as did the outer membrane activities both decrease. However, there is considerable controversy (69) and some unexplainable differences, which make the thyromimetric basis of the effects of clofibrate difficult to support.

The research discussed in this paper thus focuses attention on the influence of clofibrate on liver mitochondria. It attempts to rationalize the specific ability of the drug to increase net levels of mitochondria in terms of an inhibitory action on mitochondrial neutral protease. The over-all concepts of how the increased amount of liver mitochondria might be related to the hypo- cholesterolemic and hypotriglyceridemic effects of clofibrate were outlined earlier, in the general introduction.

Acknowledgments-The authors are grateful to Dr. Jerome Noble of Ayerst Laboratories for kindly providing samples of clofibrate. In addition, the skilled technical assistance of Mr. John T. Spears is much appreciated, as is the help of Mr. Steve Bryant for some preliminary studies.

REFERENCES

1. THORP, J. M., AND WARING, W. S. (1962) Nature 194,948 2. OLIVER, M. F. (1962) Lancet 1, 1329 3. GRUNDY,S.M.,AHEENE,E.H.,SALEN,G.,SCHREIBMAN,P.H.,

AND NISTEL, P. J. (1972) J. Lipid Res. IS, 531 4. CHENG, C. Y.. AND FELDMAN. E. B. (1971) Biochem. Pharmacol.

20,3&09 ’ . .

5. GOULD, R. G., SWYRYD, E. A., AVOY, D., AND COAN, B. (1967) Prow. Biochem. Phannacol. 2.345

6. OLIVER, M. F. (1967) Progr. Biochem. Pharmacol. 2,315 7. HESS, R., RIESS, W., AND STHUBLI, W. (1967) Progr. Biochem.

Pharmucol. 2, 325 8. BEST, M. M., AND DUNCAN, C. H. (1964) J. Lab. Cl&. Med. 64.

634 9. HESS, R., STXUBLI, W., AND RIESS, W. (1965) Nature 208,856

10. PAGET, G. E. (1963) J. Atheroscler. Res. 3, 729 11. KURUP, C. K. R., AITHAL, H. N., AND RAMASARYA, T. (1970)

Biochem. J. 116, 773 12. KRISHNAKANTHA. T. P., AND KURIJP. C. K. R. (1972) B&hem.

J. 130, 167 . ’ .

13. WHITEHOUSE, M. W., STAPLE, E., AND GURIN, S. (1961) J. Biol. Chem. 236, 68

14. HOLLOWAY, P. W..(1970) Lipid Metabolism 1,371 15. KRITCHEVSKY, D., TEPPER, S. A., SALLATA, P., KABAKJIAN,

* C. Hoppel, personal communication.

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

J.e~.;~,~~~~~~~, V. J. (1969) Proc. Sot. Exp. Biol. 43. KRISANAKANTHA, T. P., AND KURUP, C. H. R. (1973) B&&m.

16. HORN~‘NO, h. G., HERBERT, R. M., ROTH, R. J., DAVIS, D. L., Pharmacol. 22, 1245

44. GEAR, A. R. L. (1970) Biochem. J. 129,577 HORNING, E. C., FISCHER, E. P., AND JOFAN, G. L. (1972) 45. GEAR, A. R. L. (1965) D.Phil. Dissertation, Oxford University, Lipids 7, 115 Oxford, England

17. AVOY. D. R.. SWYRYD. E. A.. AND GOULD, R. G. (1965) J. 46. NEUBERT. D., OBERDISSE, E., AND BASS, R. (1968) In Bio- Lip& Res. 6, 369 .

18. BURCH, R. E., AND CURRAN, G. L. (1969) J. Lipid Res. 10,668 19. CLINRENBEARD. K. D.. SIJGIYAMA, T., Moss, J., REED, W. D.,

AND LANE, M: D. (lb73) J. Bioi. Chem. 24b, 2275 20. FALLON, J. J., ADAMS, L. R., AND LAMB, R. G. (1972) Lipids 7,

106 21. GEAR, A. R. L., AND BEDNAREK, J. M. (1972) J. Cell Biol. 64,

325 22. GEAR, A. R. L. (1965) Biochem. J. 96,118 23. SAWANT, P. L., SHIBKO, S., KUMTA, U. S., AND TAPPEL, A. L.

(1964) Biochim. Biophys. Acta 86, 82 24. OCHOA, S. (1955) Methods Enzymol. 1,735-739 25. SOTTOCASA, G. L., KUYLENSTIERNA, B., ERNSTER, L., AND

BERGSTRARD, A. (1967) J. Cell Biol. 32,415 26. HAYAISHI, 0. (1962) Methods Enzymol. 5,807-W 27. TABOR, C. W., TABOR, H., AND ROSENTHAL, S. M. (1954) J.

Biol. Chem. 208, 645 28. LINAARDT, K., AND WALTER, K. (1963) In Method8 in Enzymatic

Ana&sis (BERGMEYER, H. V., ed) pp. 783-785, Academic Press, New York

29. ALBERTI, K. G. M. M., AND BARTLEY, W. (1963) Biochem. J. 87,104

39. GOOD, N. E., WINGET, G. D., WINTER, W., CONNOLLY, T. M., ISAWA, S., AND SINGH, R. M. M. (1966) Biochemistry 6,467

31. GEAR, A. R. L., AND LEHNINGER, A. L. (1968) J. Biol. Chem. 242,3953

32. WILLIAMS, J. N. (1964) Arch. Biochem. Biophys. 107, 537 33. VANNESTE, W. H. (1966) B&him. Biophys. Acta 113,175 34. SCHNEIDER, W. C. (1945) J. Biol. Chem. 161,293 35. BURTON, K. (1956) Biochem. J. 62,315 36. ARIAS, I., DOYLE, D., AND SCEIMKE, R. T. (1969) J. Biol.

them. 144,3303 37. BEATTIE, D. S., BASFORD, R. E., AND KORITZ, S. B. (1966)

47.

48.

49. 50.

51.

52. 53. 54.

55.

56.

57.

58.

59. 60. 61. 62.

63.

chemical A8peCt8 of the Biogenesis of Mitochondria (SLATER, E. C., TAGER, J. M., PAPA, S., AND QUAGLIARIELLO, E., eds) pp. 103131, Adriatica Editrice, Bari

KROON, A, M., AND AAIJ, C. (1968) In Biochemical Aspects of the Bioaenesis of Mitochondria @LATER. E. C.. TAGER. J. M.. PAPA, ‘s., AND &UAGLIARIELL~, E., ed$) pp.‘207-228, Adri: Utica Editrice, Bari

BEDNAREK, M. M. (1973) Ph.D. Dissertation, The University of Virginia, Charlottesville, Va.

GEM, A. R. L. (1971) Fed. Proc. So,1225 BEDNAREE, M. M., AND GEAR, A. R. L. (1972) Fed. Proc. 31,

3519 SWICK, R. W., REXROTHE, A. K., AND STANGE, J. L. (1968) J.

Biol. Chem. 242,358l BEATTIE, D. S. (1969) B&hem. Biophy8. Re8. Commun. 25.721 BEATTIE, D. S. (1969) Biochem. Biophys. Re8. Commun. 36,67 BEATTIE, D. S., PATTON, G. M., AND STUCHELL, R. N. (1970)

J. Biol. Chem. 246.2177 BRAUN, G. A., MARSH, J. B., AND DRABKIN, D. L. (1963)

Biochim. Biophys. Acta 72, 647 ALBERTI, K. G. M. M., AND BARTLEY, W. (1965) Biochem. J.

96,641 ALBERTI, K. G. M. M., AND BARNEY, W. (1969) B&hem. J.

111. 763 BARTLEY, W., AND BIRT, L. M. (1970) In hsay8 in Cell Metabo-

lism. DD. 1-44. Wilev-Interscience, New York , .- TRUMAN, D. E. 6. (1983) Ezp. Cell dd8.81,313 MASON, T. L., AND SCHATZ, G. (1973) J. Bill. C&m. 243.1355 TZAGOLOFF, A. (1971) J. Biol. Chem. 246,3&O NEUPERT, W., BBDICZKA, D., AND BUCHER, TH. (1967) B&hem.

Biophys. Res. Commun. 27,488 BURKE, J. P., AND BEATTIE, D. S. (1973) Biochem. Biophys.

Res. t%mmun. 61, 349 Biochemistry 6, 926 38. BEATTIE, D. S., BASFORD, R. E., AND KORITZ, S. B. (1967) 64. PANINI, S. R. (1972) Biochem. J. 128,5OP

J. Biol. Ch-em. 242,3366 65. KATYAL, S. L., SAHA, J., AND KABARA, J. J. (1972) Biochem. 39. BEATTIE, D. S., AND IBRAHIY, N. G. (1973) Biochemistry 12, Pharmacol. 21, 747

177 66. MACKERER, C. R., HAETTING)ER, J. R., AND HUTSELL, T. C. 40. ROODYN, D. B., REISS, P. J,, AND WORK, T. S. (1961) Biochem. (1973) Biochem. Pharmacol. 22,513

J. 39,9 67. EHNHOLM, C., AND ZILVERSYIT, D. B. (1973) J. Biol. Chem. 41. MURPHY, J. B., AND KIES, M. W. (1960) Biochim. Biophys. 248, 1719

Acta 46,332 68. OKAMOTO, H. (1971) Biochem. Biophys. Re8. Commun. 43,327 42. LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL, 69. WESTERFELD, W. W., ELWOOD, J. C., AND RICHERT, D. A.

R. J. (1951) J. Biol. Chem. 193,265 (1972) Biochem. Pharmacol. 21.1117

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Adrian R. L. Gear, Arlene D. Albert and Jana M. BednarekBiogenesis: A ROLE FOR NEUTRAL MITOCHONDRIAL PROTEASES

The Effect of the Hypocholesterolemic Drug Clofibrate on Liver Mitochondrial

1974, 249:6495-6504.J. Biol. Chem. 

  http://www.jbc.org/content/249/20/6495Access 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/249/20/6495.full.html#ref-list-1

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

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from