See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/22513859

Rat liver peroxisomes catalyze the B-oxidation of fatty acids

Article  in  Journal of Biological Chemistry · April 1978

Source: PubMed

CITATIONS READS
385 1,069

1 author:

Paul B Lazarow
Institut Pasteur
121 PUBLICATIONS   10,379 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Paul B Lazarow on 23 October 2015.

The user has requested enhancement of the downloaded file.

 THE JOURNAL 01‘ B~WGICAL Cmmsmv
Vol. 253, No. 5, Issue of March 10, pp. 1522-1528, 1978
Printed m U.S.A.

Rat Liver Peroxisomes Catalyze the /3 Oxidation of Fatty

Acids*
(Received for publication, June 22, 1977)

PAUL B. LAZAROW
From The Rockefeller University, New York, New York 10021

Peroxisomes were purified by differential and equilibrium

Downloaded from http://www.jbc.org/ at Rockefeller University Library on October 23, 2015

studies of mitochondrial /? oxidation were used to investigate
density centrifugation from the livers of rats treated with whether or not purified peroxisomes are capable of carrying
clofibrate to enhance their peroxisomal system of fatty acid out the specific reactions of /3 oxidation. The appropriate 4-
oxidation. These purified peroxisomes were tested for the carbon acyl-CoA substrates (butyryl-CoA, crotonyl-CoA, and
presence of crotonase, /3-hydroxybutyryl-CoA dehydrogen- acetoacetyl-Cob) were employed, and the products were char-
ase and thiolase using spectroscopic techniques that utilize acterized by spectroscopic and enzymatic means. The suhcel-
the characteristic absorption bands of the appropriate 4- lular distribution of the enzymes of p oxidation was investi-
carbon acyl-CoA substrates. All three enzymes were found. gated with the techniques of analytical cell fractionation
Analysis of the fractions from equilibrium density centrifu- developed by de Duve and his collaborators (4). The results
gation revealed major peaks of these enzyme activities in demonstrate that peroxisomes (as well as mitochondria) con-
peroxisomes and excluded contamination by mitochondria tain crotonase, /3-hydroxybutyryl-CoA dehydrogenase, and
as an explanation of the results. In the presence of excess thiolase activity. Furthermore, peroxisomes catalyze the pro-
CoA the purified peroxisomes oxidized palmitoyl-CoA to duction of acetyl-CoA from palmitoyl-CoA. Therefore the
acetyl-CoA, and reduced NAD, with a 1:5:5 stoichiometry. peroxisomal fatty acid oxidation is a j3 oxidative process. The
The peroxisomes were inactive with butyryl-CoA and less results obtained thus far suggest that the peroxisomes are a
active with octanoyl-CoA than with lauroyl-CoA or palmit- major site for the /3 oxidation of long chain fatty acids.’
oyl-CoA; they appear specialized for the /3 oxidation of long
MATERIALS AND METHODS
chain fatty acids.
The spectroscopic assays we have used are based on those de-
scribed by Lynen and Ochoa (2) and by Mahler (3). They make use of
the characteristic absorption bands of the substrates and products of
It is generally believed that the p oxidation of fatty acids the reactions under investigation (Fig. 1). All of the acyl-CoA
occurs in mitochondria in mammalian cells. The enzymes of compounds of Fig. 1 absorb at 260 nm, due to the CoA itself, and in
the vicinity of 233 nm, hue to the thioester linkage. In addition,
/3 oxidation (acyl-CoA dehydrogenases, enoyl-CoA hydratase
crotonyl-CoA and acetoacetyl-CoA have their own characteristic
or “crotonase,” P-hydroxyacyl-CoA dehydrogenase, and thiol- absorption bands described below. A Cary 14, a Gilford 220 and a
ase) are thought to be localized exclusively in mitochondria. Gary 118 spectrophotometer were used in various parts of this
The results presented in this paper demonstrate that this is investigation.
not the case. These enzymes are present in peroxisomes as Crotonase was assayed by measuring the disappearance upon
hydration (Fig. 1) of crotonyl-CoA’s specific absorption band at 263
well, where they catalyze the p oxidation of long chain fatty nm. This absorption at 263 depends on both the double bond and the
acids. thioester linkage (2). The reaction mixture consisted of 50 nmol of
Recently Lazarow and de Duve (1) found that rat liver crotonyl-CoA and 0.04% Triton X-100 in 1 ml of 25 rnM Tris/HCl
peroxisomes are capable of oxidizing palmitoyl-Cob, with the buffer, pH 8.0, at 25”, unless indicated to the contrary in the figure
legends. The change in A,,, was converted to nanomoles of crotonyl-
reduction of O2 to Hz02 and NAD to NADH. In the presence of
CoA hydrated using AE = 6.4 mM-’ cm-’ (5).
CoA, several moles of NAD were reduced per mole of palmit- P-Hydroxybutyryl-CoA dehydrogenase was assayed in the reverse
oyl-CoA added, and the reactions were uninhibited by 1 mM direction from that shown in Fig. 1, by following the oxidation of
KCN. These data were consistent with a pathway of /3 NADH at 340 nm with acetoacetyl-CoA as substrate. The l-ml
oxidation, but in those initial studies we did not investigate reaction mixture contained 50 PM acetoacetyl-CoA, 55 PM NADH,
0.01% Triton X-100, 12 mM dithiothreitol, 1 rnM KCN, and 25 rnM
whether the site of oxidation was indeed the p carbon, nor TrislHCl buffer, pH 8.0, at 25”. Ac = 6.22 rnM-’ cm-’ (5). Under these
whether acetyl-CoA was in fact, the product. Therefore no conditions there was some nonenzymatic hydrolysis of the substrate;
conclusion was drawn concerning the exact mechanism. subsequently I found that this may be alleviated by lowering the
In the present study, the spectroscopic techniques developed dithiothreitol concentration to 0.1 mM.
Thiolase was assayed by measuring the specific absorption band
by Lynen and Ochoa (2) and by Mahler (3) in their classical
at 233 nm due to the thioester bond. Acetoacetyl-CoA was used as
* This research was supported by National Science Foundation substrate and the increase inA,,, was followed as 1 nmol of substrate
Grant PCM76-16657 and by National Institutes of Health Grants was converted to 2 nmol of acetyl-CoA. The reaction mixtures
AM19394, HL20909, and RR07065. The costs of publication of this
article were defrayed in part by the payment of page charges. This ’ These results were presented as a poster at the 11th FEBS
article must therefore be hereby marked “advertisement” in accord- Meeting in Copenhagen, August 14-19, 1977 (Lazarow, P. B. (1977)
ance with 18 U.S.C. Section 1734 solely to indicate this fact. 11th FEBS Meeting, Abstract L-A-117).

1522
 Rat Liver Peroxisomes Catalyze Fatty Acid p Oxidation 1523
usually contained 50 nmol of acetoacetyl-CoA and 100 nmol of CoA When this method was used to assay acetyl-CoA produced in
in 1 ml of 2.5 rn~ Tris/HCl buffer, pH 8.0 at 25”. AC = 4.5 IIIM-’ cm-’ reactions using peroxisomes, we first precipitated the peroxisomal
(5). proteins with cold perchloric acid, neutralized with KOH and
In some experiments acetoacetyl-CoA was detected by measuring KHCO,, and centrifuged down the KClO, and precipitated proteins.
a difference spectrum in the presence and absence of 10 mM MgCl,. When known amounts of acetyl-CoA were carried through this
Acetoacetyl-CoA in Tris buffer, pH 8, has a small absorbance band procedure the assay was still linear, but on some occasions the
at 303 nm (Fig. 2, lower curue) due to formation of the enolate ion recovery was less than 100% (Fig. 3, open circles).
(pK = 8.5; Ref. 2). The intensity of this band is greatly increased by It appears that the acetyl-CoA is not actually lost, but rather is
the addition of MgZ+ ions (Fig. 2, upper curue and inset), which underestimated in the assay due to incompletely removed perchlo-
probably form a chelate structure (2). Some attempts were made to rate making the enzyme reactions sluggish. Perchlorate precipita-
make use of this effect in the routine assay of thiolase, but MgCl, tion may be incomplete if the pH becomes slightly alkaline rather
was found to inhibit the reaction. than remaining just under 7 during neutralization or if the samples
Acetyl-CoA was assayed by means of the citrate synthetase are not kept ice-cold during centrifugation. As shown in the inset in
reaction coupled to the malate dehydrogenase reaction, essentially Fig. 3, the supernatant after perchloric acid precipitation may
as described by Decker (6). The assay conditions were slightly interfere with the determination of acetyl-CoA assayed directly. In
modified to increase the sensitivity: we used a reaction volume of 1 the experiments described in this paper, acetyl-CoA standards were
ml containing 0.5 rnM nn-malate, 0.15 rnM NAD, and 0.2 M Tris/HCl, always included and carried through the perchloric acid treatment,
pH 8.1, at 25”. The sequential increases in A,,, upon addition of 0.02 and in several cases, known amounts of acetyl-CoA were added to
unit of malate dehydrogenase and 0.17 unit of citrate synthetase duplicate experimental aliquots.
were measured in a Cary 118 spectrophotometer. They served to The peroxisomes used in these experiments were purified by
calculate the amount of acetyl-CoA present, taking into account the differential and equilibrium density centrifugation from the livers

Downloaded from http://www.jbc.org/ at Rockefeller University Library on October 23, 2015

nonlinear stoichiometry between acetyl-CoA reacted and NAD re- of three rats treated for 2 weeks with 0.5% cloiibrate in their food, as
duced (6). This method was found to be accurate and reproducible described previously (1). Cloilbrate was used because it increases
when known amounts of acetyl-CoA were added directly to the assay the activity of the peroxisomal system of fatty acid oxidation by
mixture (solid circles in Fig. 31, and the assay is believed to be approximately 1 order of magnitude (1). Based on analysis of the
highly specific for acetyl-CoA (6). marker enzymes, these peroxisomes were about 80% pure (data of
Fig. 15, Table IV, and Ref. 7). The principal contaminant on a
protein basis was microsomes, with smaller amounts of lysosomes
Ctt-CH2-CH2-C-SCoA
fl and mitochondria. The enzymes under investigation were assayed
in all the fractions from the equilibrium density centrifugation in
2H Bufyryl-CoA dehydrogenase order to verify that the enzyme activities measured in “purified
I-- peroxisomes” belonged to the peroxisomes themselves, and not to
0 any of the contaminants.
CH3-CH=CH-!-SCoA Protein, catalase, cytochrome oxidase, glucose-6-phosphatase,
and acid phosphatase were assayed according to Leighton et al. (7).
W’ Crotonose A &263 = -6.4 Palmitoyl-CoA oxidation was assayed as described by Lazarow and
l- de Duve (see legend to Fig. 6 of Ref. 1). Carnitine acetyltransferase
?H B
was measured essentially according to Solberg et al. (8).
CH3-CH-CH2-C-SCoA The following units are employed: protein (Lowry) is expressed as
NAD+
milligrams based on a bovine serum albumin standard. Catalase
p- hydroxybutyryl-Co8
and cytochrome oxidase obey first order reaction kinetics: one unit
NADH + H+ dehydrogenose At& = 6.22 of catalase decreases the H,O, concentration lo-fold/min at 0” in a
F
volume of 50 ml (7); 1 unit of cytochrome oxidase oxidizes 90% of the
8 e
CH3- C - CHZ - C - SCoA cytochrome c/min at 25” in a volume of 100 ml (9). All other enzyme
units are in micromoles/min. In the case of the palmitoyl-CoA
CoASH Thlolase A&233=45 oxidation assay this refers to micromoles of NAD reduced.
l-
Acyl-CoAs were purchased from Sigma (St. Louis) or P-L Bio-
0 chemicals (Milwaukee). Their concentrations were calculated by
2 CH,-%-SCoA dividing the absorbance at 260 by a millimolar extinction coefficient
of 16 (5) (or 22.6 for crotonyl-CoA) and multiplying by the stated
FIG. 1. Enzyme reactions investigated and specific absorption
purity of the manufacturer which varied between 79 and 91%.
bands used. Changes in millimolar extinction coefficients are given.
Cloiibrate was generously provided by Ayerst Laboratories (New
SCoA, coenzyme A in thioester linkage.
York).

o- MgCI2 fmMl
A

04.
~ +tdg++

- btg++
0.2 -

PI L.
200 250 300 350 400 Acetyl-CoA added lnmol)
Wavelength (nm) FIG. 3. Assay of acetyl-CoA directly (0) and after perchloric acid
FIG. 2. Effect of 10 mM MgCl, on the absorption spectrum of U’CA) precipitation (0). The effect of added neutralized perchloric
acetoacetyl-CoA. The inset shows the effect of the Mg*+ concentra- acid supernatant (without acetyl-CoA) on the determination of 10
tion on the magnitude of the A,,,. nmol of acetyl-CoA measured directly is shown in the inset.
1524 Rat Liver Peroxisomes
RESULTS
Crotonase - Peroxisomes catalyze a decrease in the absorp-
tion characteristic of crotonyl-CoA (Fig. 4). The reaction stops
with 21% of the crotonyl-CoA remaining, indicating
hydration, not hydrolysis, is the mechanism. This stoichiom-
etry agrees with the value of 77% hydration that may be
calculated from the equilibrium constant of 0.062 reported by
Stern (10).
As illustrated in Fig. 5, the rate of the crotonase reaction is
considerably increased by the inclusion of Triton X-100 in the
reaction mixture. The rate of the reaction is proportional
the amount of peroxisomal protein added, both in the presence
and absence of 0.04% Triton X-100 (Fig. 6). In the former case,
the slope of the regression line fitted by least squares analysis
gives a specific activity of 27 ymol of crotonyl-CoA hydrated/
min/mg of peroxisomal protein. The specific activity is 10
times lower in the absence of the detergent.
Catalyze Fatty Acid p Oxidation
Linked Sequential Reactions-The results described thus
far indicate that peroxisomes are capable of catalyzing each of
the crotonase, P-hydroxybutyryl-CoA dehydrogenase, and thi-
that olase reactions. This conclusion was tested further by attempt-
ing to link the three reactions to see whether the product of
the first would serve as the substrate of the second, etc.
In the experiment illustrated in Fig. 13, 33 nmol of crotonyl-
CoA were placed in a cuvette in 1 ml of 50 mM Tris/HCl buffer,
pH 9.0, and the spectrum was measured (Curue A). Peroxi-
somes were added to both sample and reference cuvettes, and
to the spectrum was remeasured (Curve B). There was a drop in
the absorbance at 263 nm corresponding to the hydration of 28
nmol of crotonyl-CoA. NAD was then added to both cuvettes.
Small absorption peaks appeared at 340 and 303 nm (Spectrum
C), corresponding to the reduction of 5 nmol of NAD to NADH
and the formation of 5 nmol of acetoacetyl-CoA. The reaction
was incomplete; from the equilibrium constant of 6.3 x lo-”
given by Wakil (11) one would expect to form 8 nmol of

Downloaded from http://www.jbc.org/ at Rockefeller University Library on October 23, 2015

PHydroxybutyryl-CoA Dehydrogenase - As shown in Fig.
7, peroxisomes catalyze the oxidation of NADH with acetoace- acetoacetyl-Cob. At this point CoA was added to both cu-
tyl-CoA as substrate. The rate of the reaction is proportional vettes: formation of acetyl-CoA would make the overall path-
to the amount of peroxisomal protein added (Fig. 8) and way exergonic. As illustrated in Curve D, the absorption peak
corresponds to a specific activity of 1.4 pmol/min/mg of perox- at 340 nm increased, corresponding to the reduction of a
isomal protein. further 10 nmol of NAD to NADH. The peak at 303 nm
Thiolase - Peroxisomes catalyze an increase in thioester vanished, indicating the disappearance of the acetoacetyl-
bond absorbance (Fig. 9); no reaction occurs if either acetoace- CoA. The absorbance at 233 increased by an amount corre-
tyl-CoA or CoA is omitted. The observed increase of 31 nmol sponding to the formation of 14 nmol of thioester bond,
of thioester linkage is quantitatively consistent with the presumably by conversion of 14 nmol of acetoacetyl-CoA to 28
expected doubling of acyl-CoA. In a duplicate reaction an
increase of 32 nmol was observed.
The spectra measured in a subsequent experiment (Fig. 10)
demonstrate both a decrease in the size of the small absorption
band at 303 nm, characteristic of acetoacetyl-CoA, and the
increase at 233 nm due to thioester bond formation. The AA340 ffyy--J * ‘:j,7
disappearance of substrate during the reaction is shown more
clearly by the Mg*+ difference spectra of Fig. 11. The formation
of acetyl-CoA is given in Table I, together with the stoichiom-
etry of the other changes observed. It is clear from these data 0 2 4 0 2 4 6
that there was a nearly quantitative conversion of acetoacetyl- Minutes Peroxisomal
CoA to acetyl-CoA in a 1:2 ratio. protein (pg)

Fig. 12 illustrates that the rate of the reaction depends FIG. 7 (left). Time course of /3-hydroxybutyryl-CoA dehydrogen-
linearly on the amount of peroxisomes added. The specific ase reaction, assayed in reverse. The reaction was initiated by the
addition of 6.9 +g of peroxisomal protein.
activity is 2.8 pmol of acetoacetatelminlmg of peroxisomal FIG. 8 (right). Effect of peroxisome concentration on the rate of
protein. the fi-hydroxybutyryl-CoA dehydrogenase reaction.

nmol
xii

-01
AA263

0+ 4
Minutes Triton X-100(%)

FIG. 4 (left). Time course of the hydration of crotonyl-CoA cata- rate of the crotonase reaction. Each reaction mixture contained 40
lyzed by 0.7 (A) or 1.4 pg (B) of peroxisomal protein. The base-line nmol of crotonyl-CoA and 1.4 pg of peroxisomal protein.
is drawn at -0.245 A, the decrease observed upon alkaline hydroly- FIG. 6 (right). Effect of peroxisome concentration on the rate of
sis of an identical aliquot of crotonyl-CoA, from which we calculate the crotonase reaction in the presence (x) and absence (0) of 0.04%
that there were 38 nmol of substrate present. Triton X-100.
FIG. 5 (center). Effect of the Triton X-100 concentration on the
 Rat Liver Peroxisomes Catalyze Fatty Acid j3 Oxidation 1525
O-

5-

A 05 o-

AA

010
AA 233 02 5-

005

250 300

Downloaded from http://www.jbc.org/ at Rockefeller University Library on October 23, 2015

Mtnules Wavelength lnmi Wovelength inm)

FIG. 9 (left). Time course of the thiolase reaction. The reaction somal protein/ml. The reaction was followed at 233 nm; when it was
mixture contained 30 nmol of acetoacetyl-CoA, 100 nmol of CoA, and complete, the second spectrum was taken.
17 pg of peroxisomal protein in 1 ml of 30 mM phosphate buffer, pH FIG. 11 (right). Mg*+ difference spectra before and after the
7.4, at 37”. thiolase reaction to test for acetoacetyl-CoA. Duplicate aliquots of
FIG. 10 (center). Spectra before and after the thiolase reaction. the reaction mixture (before addition of peroxisomes) were placed in
The sample cuvette contained 47 PM acetoacetyl-CoA, 100 /AM CoA, sample and reference cuvettes. Their difference spectrum was every-
and 25 rnM Tris/HCl buffer, pH 3.0. The reference cuvette contained where zero as expected. Then 10 rnM MgClz was added to the
CoA and Tris/HCl only. After measuring the spectrum, and remov- sample cuvette, 1 mM EDTA was added to the reference cuvette and
ing aliquots for later determination of acetyl-CoA and a MgZ+ the difference spectrum was measured (top curve). A similar differ-
difference spectrum, purified peroxisomes were added to both sample ence spectrum was measured on duplicate aliquots of the reaction
and reference cuvettes at a final concentration of 6.9 pg of peroxi- mixture after the thiolase reaction was over (bottom curve).

TABLE I
of thiolase reaction
Stoichiometry
The reaction is described in the legend to Fig. 10. No acetyl-CoA
was detected in the reaction mixture before addition of peroxisomes
nor in the peroxisomes themselves.
Component nmol/ml
m
Acetoacetyl-CoA added 47 min
Acetoacetyl-CoA after reaction 2
Thioester bond, net increase 44
Acetyl-CoA formed 91

nmol of acetyl-CoA. Spectrum D’ was taken 30 min later,

when the reactions were complete; there was a net increase of I , b

22 nmol of thioester bond and a reduction of a total of 26 nmol 5 IO 15

of NAD. Peroxisomol protein (pg)

In order to show the transient accumulation of acetoacetyl- FIG. 12. Effect of peroxisome concentration on the rate of the
thioiase reaction. Reaction conditions as in Fig. 10.
CoA more clearly, this experiment was repeated exactly as
described, except that 10 times more NAD was added in order
to shift the @-hydroxybutyryl-CoAJacetoacetyl-CoA equilib- alkaline pH of 9 that was employed. Hydrolysis may be
rium forward. This, however, prevented measuring spectra reduced by working at pH 8, but this shfis the equilibrium of
below 280 nm, due to the intense absorption of the NAD. As the /3-hydroxybutyryl-CoA dehydrogenase reaction further
illustrated in the lower part of Fig. 13, Spectrum C, use of away from acetoacetyl-CoA.
this larger amount of NAD considerably increased the sire of Butyryl-CoA Dehydrogenase - Since crotonyl-CoA serves as
the absorption peaks at 303 and 340 nm. They correspond to substrate for the linked reactions leading to acetyl-CoA, we
the accumulation of 11 nmol of acetoacetyl-CoA and the made use of this fact to assay for the desaturation of butyryl-
reduction of 12 nmol of NAD, respectively. Addition of CoA CoA (Fig. 1, first reaction). Butyryl-CoA was added to a
again resulted in the disappearance of the acetoacetyl-CoA reaction mixture containing peroxisomes, NAD, and CoA
and further reduction of the NAD. As expected, the total (Table II) and the formation of NADH was measured spectro-
NADH formation (29 nmol, Curve D’) was similar to that photometrically at 340 nm. No reaction was detected. Addition
observed in the first experiment. of crotonyl-CoA demonstrated that the assay system was
These experiments were repeated several times, taking a functioning correctly.
number of spectra at each step, and generally similar results Oxidation ofPalmitoyl-CoA to Acetyl-CoA -A l-ml reaction
were obtained. However, the final yield of thioester bond mixture was prepared containing 10 nmol of palmitoyl-CoA
formed is lower when many spectra are taken at step C, due together with excess NAD and CoA. The addition of peroxi-
to gradual hydrolysis of acetoacetyl-CoA which occurs at the somes resulted in the reduction of 52 nmol of NAD to NADH
1526 Rat Liver Peroxisomes Catalyze Fatty Acid /3 Oxidation
TABLE II
Coupled assay for butyryl-CoA dehydrogenase
Reactions contained 10 nmol of substrate and 8.6 pg of peroxi-
somal nrotein: other conditions as in Table III.
NAD reduction
Substrate
A Crotonyl-CoA Initial rate Extent
B +Peroxlsomes nmollmin n?nol
C +NAD Butyryl-CoA 0 0
D +CoA Crotonyl-CoA 4.0 8.5

TABLE III
Acetyl-CoA formed by peroxisomes from palmitoyl-CoA
Palmito l- Ace;&C~A
Reaction mixture” CoA ad B ed “Eel-
nmol nmol ?lOZOl

Complete
Experiment 1 10 52 50 2 1

Downloaded from http://www.jbc.org/ at Rockefeller University Library on October 23, 2015

0 0 0
Experiment 2 10 48 48 + 2
0 0 0
Peroxisomes 10 0 0
omitted
D Complete reaction mixtures contained 100 nmol of CoA, 200
nmol of NAD, and 34 pg of peroxisomal protein in 1.00 ml of 50 mM
TrislHCl buffer, pH 8.0, with 1 mM dithiothreitol, 0.01% Triton X-
100 and 0.0075% bovine serum albumin, at 37”.
* Acetyl-CoA was determined on quadruplicate aliquots of each
reaction mixture and corrected for recovery of standard acetyl-CoA
added to half of the aliquots; the recovery was 75% in the first
experiment and 101% in the second. Means and standard deviations
250
are indicated.
Wavelength (nm)
FIG. 13. Sequential p oxidation reactions. Upper graph: A, spec-
trum of crotonyl-CoA (33 nmol by alkaline hydrolysis) in 1 ml of 50
mu Tris/HCl buffer, pH 9.0, at 25”. The reference cuvette contained
buffer only; B, spectrum after addition of 34 pg of peroxisomal
protein to both sample and reference cuvettes; C, spectrum after
subsequent addition of 50 nmol of NAD to both cuvettes; D, spectrum
after subsequent addition of 50 nmol of CoA to both cuvettes; D’,
spectrum 30 min later. Lower graph: identical experiment except
that 500 nmol of NAD were added to the sample cuvette only before
Spectrum C. Curves A and B (not shown) were very similar to those
in the upper graph. These spectra were measured with a Cary 118
spectrophotometer using a constant slit width of 0.7 mm, automatic
control of the gain to maintain constant signal energy and a scan
speed of 1 rim/s. Identical spectra were observed with narrower slits
or slower scan speeds. The scans were started 4 min apart except for
D and D’ which followed their predecessors by 8 and 30 min,
respectively. Calculations of metabolite concentrations (see text)
take into account dilutions of 2 or 2.5% at each step. Chain length
FIG. 14. Ability of peroxisomes to oxidize acyl-CoAs of varying
(Table III). No NAD reduction occurred in control reaction chain lengths. Reaction mixtures contained 10 nmol of substrate and
9.6 pg of peroxisomal protein; other conditions were as described
mixtures in which palmitoyl-CoA or peroxisomes were omit- previously (l), including 0.1 mM CoA. Means and ranges of duplicate
ted. The reaction mixtures were assayed for acetyl-CoA: 50 determinations are indicated.
nmol of acetyl-CoA were formed in the 1 ml of complete
mixture with substrate (and none in the controls) (Table III). CoA and no activity was detected with butyryl-CoA as sub-
This is in good agreement with the value of 52 nmol of NAD strate. Thus the peroxisomes appear specialized for the oxida-
reduced and indicates that the 10 nmol of palmitoyl-CoA tion of longer chain length fatty acids.
went through five cycles of /3 oxidation. The experiment was Distribution of p Oxidation Enzymes after Equilibrium
repeated a second time with very similar results (Table III). Density Centrifugation - The activities of the various enzymes
Chain Length SpecifEity of Peroxisomal Fatty Acid Oxida- were assayed in all the fractions from the isopycnic centrifu-
tion - The ability of peroxisomes to oxidize fatty acyl-CoAs of gation experiment by which the purified peroxisomes used in
various chain lengths was determined by measuring the rate these studies were prepared. The results are shown in Fig.
of NAD reduction. This assay involves the concerted action of 15, together with the distributions of catalase and cytochrome
all four /3 oxidation enzymes. As shown in Fig. 14, peroxisomes oxidase, which are used as marker enzymes for peroxisomes
appear to have similar activity toward lauroyl- and palmitoyl- and mitochondria, respectively. Fig. 15 also includes the
CoA, but the activity is considerably lower with octanoyl- distributions of protein, as well as of glucose-6-phosphatase
 Rat Liver Peroxisomes Catalyze Fatty Acid /3 Oxidation 1527
TABLE IV
Crotonose Composition of peroxisome-rich fraction prepared by differential
600 centrifugation from livers ofclofibrate-treated rats
Marker for Homage- Peroxisome-rich
400 nate fraction”

200
Protein 217 9.6 4.4
Catalase Peroxisomes 125 13.7 11
Thiolose Cytochrome oxidase Mitochondria 33.5 0.93 2.8
60 1.2 6.2
Glucose-6-phospha- Endoplasmic 19.3
tase reticulum

I(
40 Acid phosphatase Lysosomes 9.5 1.3 14
a This fraction was subjected to equilibrium density centrifuga-
20 tion in the experiment illustrated in Fig. 15.
* Units are defined under “Materials and Methods.” The starting
material was 35 g of liver.
Hydroxybutyryl-CoA Density I ’
peroxisomal and mitochondrial fractions. Lastly, the distribu-

Downloaded from http://www.jbc.org/ at Rockefeller University Library on October 23, 2015

tion of palmitoyl-CoA oxidizing activity (measured as the
rate of NAD reduction (1)) is included. This appears to be
virtually identical with that of catalase; no additional activity
in the region of the mitochondria is evident.
The peak peroxisomal fraction of Fig. 15 was used in all the
other experiments described previously.
Cotolose / 20 DISCUSSION

The results demonstrate that hepatic peroxisomes of clofi-

60
brate-treated rats contain crotonase, p-hydroxybutyryl-CoA
40 Acid phosphotose dehydrogenase, and thiolase. Furthermore, the peroxisomes
convert palmitoyl-CoA to acetyl-CoA. Therefore, the peroxi-
(lysosomes) /
20 somal fatty acid-oxidizing system reported previously (1) pro-
h ceeds by a mechanism of p oxidation.
It should be emphasized that although these experiments
Cytochrome
oxldose were performed on peroxisomes from rats that had been
(mrtochondriol 1 treated with cloiibrate in order to increase the activity of
Glucose-6-phosphotose their peroxisomal system of fatty acid oxidation (11, hepatic
(microsomes) I peroxisomes from normal rata also contain these three en-
zymes, although at much lower levels.Z Thus clotibrate affects
the activity, but not the basic mechanism, of this peroxisomal
enzyme system.
oc Butyryl-CoA desaturation activity was not detected in these
purified peroxisomes. Beinert, in his review of (mitochondrial)
Volume (ml) acyl-CoA dehydrogenases (12), states that most tissues studied
FIG. 15. Distribution of enzyme activities after equilibrium den- have two (and pig liver three) acyl-CoA dehydrogenases,
sity centrifugation of the peroxisome-rich fraction described in Table active on short, (medium), and long chain acyl-CoAs. The
IV. The ordinate scales are chosen so that the areas under all the other three enzymes of p oxidation appear to be active on all
histograms are the same, in effect normalizing the distributions to
facilitate their comparison (4). The peak peroxisomal fraction was the different chain length substrates that have been tested
used for all the other experiments described in this paper. (10, 11, 13). These facts, together with the observation that
peroxisomes reduce 0, to H,O, in the presence of palmitoyl-
and acid phosphatase, marker enzymes for microsomes and CoA (1, 14) lead us to infer that peroxisomes contain an
lysosomes, respectively, which were present in small quanti- enzyme capable of the cr,p desaturation of long chain acyl-
ties in the peroxisome-rich fraction layered on the gradient CoAs (although it appears to function as an oxidase rather
(Table IV). It may be emphasized that this layer was much than a dehydrogenase), but have little or no enzyme active on
poorer in mitochondria than in peroxisomes (Table IV). short chain acyl-CoAs (which however is found in mitochon-
Crotonase, /3-hydroxybutyryl-CoA dehydrogenase, and thi- dria2).
olase each have a major peak of activity similar in shape and In our first experiments only three cycles of p oxidation
position to that of catalase, demonstrating that these enzymes were observed with palmitoyl-CoA as substrate (1). After
are present in peroxisomes. In addition, they each have a reiining the reaction conditions, five cycles have now been
shoulder or smaller peak in the center of the gradient, where detected. The apparent inactivity of the peroxisomes toward
the mitochondria are localized, as shown by the distribution short acyl-CoAs provides an explanation for the fact that we
of cytochrome oxidase. have not observed the theoretical maximum of seven cycles.
Fig. 15 also shows the distribution of carnitine acetyltrans- Acetyl-CoA was observed to accumulate as the end product
ferase, which is present in similar concentrations in the * P. B. Laxarow, unpublished results.
1528 Rat Liver Peroxisomes Catalyze Fatty Acid /3 Oxidation

of hepatic peroxisomal /3 oxidation in these experiments. This and oxidases, making them a member of the peroxisome
differs from previous studies by other investigators with family as defined by de Duve and Baudhuin (20). Blum (21)
other organelles or tissues, in which the acetyl-CoA was has reported that Tetruhymena peroxisomes contain some
further metabolized to ketone bodies, to intermediates of the enzymes of p oxidation, and Graves and Becker (22) have
Krebs or glyoxylate cycles, or to CO,. The fate of the acetyl- provided similar evidence for the glyoxysomes of Euglena.
CoA produced in the peroxisomes is not known. It may be These facts support de Duve’s speculations concerning a com-
transported to the mitochondria for further oxidation or used mon ancient evolutionary origin for animal and plant peroxi-
for biosynthetic reactions elsewhere in the cell. somes and glyoxysomes (23).
Carnitine may play a role in transporting this “active In addition to its physiological function, the hepatic peroxi-
acetate” to the mitochondria, which are believed to be im- somal system of p oxidation appears to play a role in reducing
permeable to acetyl-CoA (15). The presence of carnitine ace- serum lipid levels during therapy with hypolipidemic drugs.
tyltransferase in both peroxisomes and mitochondria was first Three different drugs greatly increase hepatic palmitoyl-CoA
reported by Markwell et al. (16). Solberg et al. (8) and Moody oxidation in the rat, at least in part by enhancing the activity
and Reddy (17) found that the activity of this enzyme is of this peroxisomal enzyme system (24).
increased in the livers of rats treated with clotibrate, such as
the rats used in these investigations. The present experiments Acknowledgments-1 would like to thank Ms. Annabella
confirm the presence of carnitine acetyltransferase in both Bushra for her expert and dedicated technical assistance, and
peroxisomes and in mitochondria. Furthermore they provide Dr. C. de Duve for his helpful criticism of this manuscript.

Downloaded from http://www.jbc.org/ at Rockefeller University Library on October 23, 2015

a plausible explanation for the presence of this enzyme in REFERENCES
peroxisomes by showing that peroxisomes actively synthesize
acetyl-CoA, the enzyme’s substrate. We infer that carnitine 1. Lazarow, P. B. & de Duve, C. (1976) Proc. Natl. Acad. Sci. U.
S. A. 73, 2043-2046
functions as an interorganellar shuttle of acetyl (and perhaps 2. Lynen, F. & Ochoa, S. (1953) Biochim. Biophys. Acta 12, 299-
other acyl) residues. 314
Until now it has been generally believed that the p oxida- 3. Mahler, H. R. (1953) Fed. Proc. 12, 694-702
tion of fatty acids is a mitochondrial function in mammalian 4. de Duve, C. (1971) J. Cell Biol. 50, 20D-55D
5. Dawson, R. M. C., Elliott, D. C., Elliott, W. H. & Jones, K. M.
cells. The observation that peroxisomes catalyze /3 oxidation (1974) Data for Biochemical Research. 2nd Ed. Oxford Univer-
raises the questions: Is there any /3 oxidation in mitochondria? sity Press, London
Was peroxisomal /3 oxidation attributed to mitochondria be- 6. Decker, K. (1974) in Methods ofEnzymatic Analvsis (Bermnever.
cause mitochondrial fractions prepared by differential centrif- H. U., ed) 2nd Ed, Vol. 4; pp. “1988-1993, “Academic Press.;
New York
ugation are contaminated by peroxisomes? The fractionation 7. Leighton. F.. Poole. B.. Beaufav. H.. Baudhuin. P.. Coffev. J.
data of Fig. 15 provide some clues. There are major peaks of w., Fowler, S. &‘de Duve, C.-(1968) J. Cell Bibl. 37, 482%13
crotonase, thiolase, and P-hydroxybutyryl-CoA dehydrogen- 8. Solberz, H. E., Aas. M. & Daae, L. N. W. (1972) Biochim.
ase in peroxisomes, and small peaks or shoulders of these Biophys. Acta 280, .434-439
enzymes where the mitochondria are located. However, since 9. Cooperstein, S. J. & Lazarow, A. (1951) J. Biol. Chem. 189, 665
670
the starting material layered on this gradient was consider- 10. Stern, J. R. (1961) in The Enzymes (Boyer, P. D., Lardy, H. &
ably enriched in peroxisomes relative to mitochondria, I Myrblck, K., eds) 2nd Ed, Vol. 5, pp. 511-529, Academic
calculate that in the cell both organelles may have roughly Press, New York
similar amounts of these enzymes. Further investigations of 11. Wakil, S. J. (1963) in The Enzymes (Boyer, P. D., Lardy, H. &
Myrbiick, K., eds) 2nd Ed, Vol. 7, pp. 97-103, Academic
their subcellular distributions are in progress. The impression Press, New York
one gets from Fig. 15 that in clofibrate-treated rats, most of 12. Beinert, H. (1963) in The Enzymes (Boyer, P. D., Lardy, H. &
the hepatic palmitoyl-CoA oxidation occurs in peroxisomes Myrbtick, K., eds) 2nd Ed, Vol. 7, pp. 447-466, Academic
has received considerable additional experimental support; Press, New York
even in normal rats, the peroxisomes make a substantial con- 13. Hartmann, G. & Lynen, F. (1961) in The Enzymes (Boyer, P.
D,, Lardy, H. & Myrblck, K., eds) 2nd Ed, Vol. 5, pp. 381-
tribution to palmitoyl-CoA oxidation.3 The data available 386, Academic Press, New York
so far suggest that both peroxisomes and mitochondria cata- 14. Lazarow, P. B. (1976)Fed. Proc. 35, 1501
lyze p oxidation, but their specificities are not the same. 15. Fritz, I. B. (1963) Adu. Lipid Res. 1, 285-334
The finding that rat liver peroxisomes oxidize fatty acids 16. Markwell, M. A. K.. McGroartv. E. J.. Bieber. L. L. & Tolbert.
provides a function of obvious physiological importance for N. E. (i973) J. B&l. Chem. zhk, 3426-3432
17. Moody, D. E. & Reddy, J. K. (1974) Res. Commun. Pathol.
the mammalian peroxisomes. Moreover, it demonstrates a Pharmacol. 9, 501-510
close functional kinship between peroxisomes in a wide vari- 18. Breidenbach, R. W. & Beevers, H. (1967)Biochem. Biophys. Res.
ety of cell types, including plants and protozoa. Ten years Commun. 27, 462-469
ago, Breidenbach and Beevers (18) reported that germinating 19. Cooper, T. G. & Beevers, H. (1969) J. Biol. Chem. 244, 3514-
3520
fatty seedlings contain an organelle with all the enzymes of 20. de Duve, C. & Baudhuin, P. (1966) Physiol. Rev. 46, 323-357
the glyoxylate cycle, which they termed the “glyoxysome.” 21. Blum, J. J. (1973) J. Protozool. 20, 688-692
Subsequently, Cooper and Beevers (19) found that this organ- 22. Graves, L. B., Jr. & Becker, W. M. (1974) J. Protozool. 21, 771-
elle catalyzes /3 oxidation. Glyoxysomes also contain catalase 774
23. de Duve, C. (1969) Ann. N. Y. Acad. Sci. 168, 369-381
3 P. B. Lazarow, manuscript in preparation. 24. Lazarow, P. B. (1977) Science 197, 580-581
 :
Rat liver peroxisomes catalyze the beta
oxidation of fatty acids.

P B Lazarow
J. Biol. Chem. 1978, 253:1522-1528.

Downloaded from http://www.jbc.org/ at Rockefeller University Library on October 23, 2015

Access the most updated version of this article at http://www.jbc.org/content/253/5/1522

Find articles, minireviews, Reflections and Classics on similar topics on the JBC Affinity
Sites .

Alerts:
• When this article is cited
• When a correction for this article is posted

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

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

http://www.jbc.org/content/253/5/1522.full.html#ref-list-1

View publication stats