in
U.S.A.
and Refeeding
and in the
THE RELATIOSSHIP
BETWEEN
REGULATION
OF KETOGENESIS
ACID
FATTY
OXIDATION
on Carbohydrate
AND
ESTERIFICATION
IN THE
J.
DENIS
m!GARRY,
J~RGEN
M.
n/IEIER,
AND
D~IEL
and Biochemistry,
The strikiiig
enhanccmcnt
in hepatic gluconcogenesis
and
ketogenesis characteristic of the fasting state is well documented.
by United
Awardee
States Public
5-K3-AM
9968,
270
June 26,1972)
FOSTERI
The
University
of Texas Southwestern
Medical
SUMMARY
* This investigation
was slipported
Health Service Grant CA-08269.
$ Research
Career Development
United States Public IIealth Service.
W.
and
271
(9). Moreover, the view that depressed triglyceride
synt,hesis
is the primary mechanism for the shunting of fatty acids into
the oxidative pathway in the ketotic state is difficult to reconcile
with the situation commonly found in experimental
diabetic
ketoacidosis, where increased hepatic concentration
of fat occurs
in the face of higher rates of ketone production than are seen in
starvation (7, IO).
The data outlined below illustrate
that surprisingly
brief
periods of starvation and refeeding exert striking effects on glucose and fatty acid metabolism in the intact rat and that a remarkable degree of synchrony exists between the events taking
place in vivo and those observed in the isolated perfused liver.
In addition, evidence will be presented to support the thesis that
in the fasted state no fundamental defect is present in the triglyceride-synthesizing
capacity of the liver, and that the marked
enhancement of ketogenesis is due to preferential flux of available
fatty acid carbon through the /? oxidation sequence, the latter
event probably resulting from the activation of an early step in
the oxidative pathway.
PROCEDURE
Animals-Male
Sprague-Dawley
rats weighing 100 to 120 g
were used in all experiments.
Animals were fed a diet containing 58.5% sucrose, 21% casein, and less than lye fat (General
Biochemicals,
Chagrin Falls, Ohio).
The term normal
as
used in the text refers to rats that had free access to food up to
the time of commencement of experiments which was generally
between 9:30 and 10:00 a.m. All fasted rats were deprived of
food for the indicated time periods. When animals were fasted
for shorter times (6 to 12 hours) they received 2 ml of a balanced
liquid diet containing
34% glucose (Diet ~~116 EC, General
Biochemicals) by stomach intubation
as a final feeding in order
to standardize the period of fasting.
Animals used in refeeding
studies received 3 ml of the liquid diet initially followed by l-ml
portions at hourly intervals until used for experiments.
Analyses on Blood and Liver from Intact Rats-Livers
and
blood samples were removed and plasma was analyzed for ketone
bodies, free fatty acids, glucose, and insulin as described earlier
(7). Liver glycogen was determined by the method of Good
et al. (11).
Liver Perfusion-Livers
were perfused with recirculating
medium using the apparatus and techniques described in previous
reports (5, 12). As routinely used in these studies aged human
erythrocytes were washed with 0.9% sodium chloride and Krebs
bicarbonate buffer, pH 7.4, before being suspended in the latter
to a hematocrit of 40%. This was mixed with an equal volume
of a solution containing
8% bovine albumin in Krebs bicarbonate buffer, pH 7.4, to provide the final perfusion fluid.
In
the present series of experiments it was observed that the initial
lactate concentration
in the cell-free fluid was regularly in the
region of 3 mM. Because of the known antiketogenic
properties
of lactate, in some experiments it was desirable to remove it
from the medium.
Since the lactate was known to originate
from the erythrocytes, this was accomplished by dialyzing the
red cells for 24 hours against 0.9% sodium chloride at 4 prior
to use.
In all of the experiments to be described, sufficient oleic acid
was included in the initial pool to give a fatty acid concentration
in the cell-free fluid of approximately
0.7 InM.
This concentration was maintained throughout
the experimental period by the
infusion of a solution containing 12.5 mM oleate and 10% albumin dissolved in 0.9% NaCl solution, pH 7.4, at a rate of 75 ~1
per min. In experiments using [1-14CJoleic acid 0.4 ml of a 5%
RESULTS
of Starvation
EXPERIMENTAL
albumin-NaCl
solution containing
20 PCi of the labeled fatty
acid (specific activity, 56.3 PCi per pmole) was injected into the
circulating fluid after 30 min of perfusion.
Thirty minutes later
livers were rapidly frozen in liquid Nz and samples of perfusate
were taken for analysis.
Analyses on Perfusion Pluid and Liver-The
concentrat,ions of
pyruvate and lactate in the cell-free fluid were determined by
the procedures described for acetoacetate and P-hydroxybutyrate, respectively
(13), with the modification
that P-hydroxybutyrate dehydrogenase was replaced by lactate dehydrogenase
(EC 1.1.1.27).
In experiments where [3-14C]lactate was included in the perfusion fluid the incorporation
of isotope into
glucose and liver glycogen was determined using slight modifications of the procedures described by Eston and Park (14).
All other methodology
has been described in previous reports
(5, 12). In the text the term ketone bodies will always refer
to the sum of acetoacetate and fl-hydrosybutyrate.
illuteri&-Unlabeled
oleic acid was obt.ained from Sigma.
In all perfusion experiments the fatty acid was bound to the
bovine albumin (Fraction V, Armour Pharmaceutical
Co., Lot
G35808) by the method of Van Harken et al. (15). [1-14C]Oleic
acid and L-[3-%]lactate
were from new England Nuclear.
272
was about 0.25 peq per ml, but this value doubled
of starvation and continued to increase until the
at which time it had reached 0.8 peq per ml. The
remained at this level throughout
the following 36
after 6 hours
12-hour point
concentration
hours but fell
E$ect
TABLE
acids
in medium
in NaCl
the
State
Normal.
Fasted12
Fasted
Fasted
Fasted
Fasted
Fasted
solution
was infused
experimental
period.
of animal
at a rate
Within
of 75 pl per min
each
group
of
Additions
.
24
48
96
24
24
hr.......................
hr.......................
hr.
hr..
hr..
.
hr and refed
.
.
3 hr. .
(+)-Decanoylcarnitine
on Metabolism
in
during
perfusion
of rat hers
with
oleic
acid
in microequivalents
per
from two to four experiments.
Minutes
ml
of
the
cell-free
per-
of perfusion
13
30
4.5
0.66
0.67
0.72
0.66
0.69
0.74
0.71
0.65
0.78
0.76
0.66
0.67
0.76
0.68
0.73
0.78
0.79
0.67
0.72
0.76
0.68
0.82
0.81
0.83
0.70
0.72
0.79
0.67
~^
60
0.85
0.85
0.82
0.74
0.74
0.85
0.76
of free fatty
and Refeeding
Perfused
Rat Liver
Ketogenesis-Previous
studies from this and other laboratories
have shown that after fasting periods of either 24 or 48 hours
the capacity of the isolated perfused rat liver to synthesize
ketone bodies from fatty acids and glucose from lactate is greatly
enhanced (2, 5, 16). Because of the rapid and pronounced
effects of starvation and refeeding on the metabolic parameters
studied in viva (Fig. 1) the possibility existed that similar kinet.ics
might be observed with regard to changes in fatty acid and
glucose metabolism in the isolated liver. To this end a second
batch of rats was subjected to periods of fasting and refeeding
identical with those used in the in viva experiments.
At the
appropriate
times the livers were perfused with oleic acid as
described under Experimental
Procedure.
Since our primary
interest during this investigation was in the effects of starvation
and refeeding on a hepatic fatty acid metabolism it was decided
to choose an initial concentration
of fatty acid close to that observed in the plasma of fasted rats and to maintain this level in
To
the perfusion fluid throughout the course of the experiment.
illustrate that these conditions were in general satisfied, representative examples showing the fatty acid concentration
in the
perfusion fluid at different time points are shown in Table I.
It is seen that, irrespective of the nutritional
state of the donor
animal, the uptake of oleate by the liver was approximately
the
same and remained fairly constant with time in all experimental
HOURS OF STARVATION
Concentration
of Starvation
273
-Starvation
200
8.
t
IO
,,
15 20
25
30 Gk+&
HOURS OF STARVATION
nondialyzed
erythrocyt.es.
The notation
glucose
production
- 35 lactate
uptake
(A)
should
be read glucose
production
minus
35 lactate
uptake.
Each point represents
the mean value
f S.E.M.
for determinations made after GO min of perfusion
with four to eight livers.
groups.
In addition, the presence of a potent inhibitor of long
chain fatty acid osidation, ( +)-decanoylcarnitine,
had no effect
on the uptake of oleate by livers from 24-hour fasted rats.
The data displayed in Fig. 2C illustrate the effects of starvation
and refeeding on ketone body production
in the perfused liver.
There is a striking qualitative similarity between the effects of
the same dietary manipulations
on ketone production
by the
isolated liver and plasma ketone levels in tivo (Fig. 10).
Thus,
little change took place in ketogenesis in the first 6 hours of
fasting.
The next 3 hours, however, were accompanied
by
clear-cut alterations in the metabolic profile of the liver.
During
this time ketone body synthesis over the 60-min perfusion period
increased sharply from 39 to 131 pmoles/lOO g of body weight.2
A maximal output of 147 pmoles/lOO g of body weight was
achieved at the 12-hour point and this was unchanged at 24
hours. Thereafter the rate declined somewhat such that at 48
and 96 hours values were in the region of 100 pmoles/lOO g of
body weight.
As was the case with starvation the effect of refeeding on ketone production
from oleate was also brisk, the
output falling from 147 to 80 ~moles/lOO g of body weight in the
first hour. By 3 hours the value was reduced to 36 pmoles/lOO
g of body weight, a rate which was not different from that exhibited by livers from nonfasted rats.
Gluconeogenesis
from Lactate-Preliminary
studies revealed
that the perfusion fluid as routinely prepared contained significant quantities of lactate (generally in the region of 3 mu).
Since this compound serves as an excellent gluconeogenic
substrate it was of interest to determine the effects of starvation
and refeeding on its uptake by the liver. These data are shown
HOURS OF STARVATION
30
S.E.M.
perfusion
four
with
for determinations
livers.
in Fig. 2B. As was the case for ketone production, the rate of
lactate uptake increased between 6 and 9 hours of starvation
and plateaued between 12 and 24 hours, at which time an uptake
of 212 pmoles/lOO g of body weight was recorded.
After 48
hours this value had fallen to 177 and remained at this level
during a further 2 days of fasting.
It is noteworthy,
however,
that while the enhanced capacity of the liver to take up lactate
from the medium and to convert oleate into ketone bodies
appeared to correlate during starvation this was clearly not the
case in the reversal phase. Thus, 2 hours of refeeding, which
caused a 54% reduction in ketogenesis, had no significant effect
on lactate uptake.
In general, therefore, the rate of decline in
lactate uptake was considerably
slower than the reversal of
ketogenesis.
In order to determine whether the amount of lactate consumed
by the livers in different nutritional
states adequately reflected
relative rates of gluconeogenesis it was necessary to repeat these
experiments using perfusion fluid containing a known quantity
of labeled lactate and to measure the incorporation
of isotope into
glucose and liver glycogen.
The results of these studies are
shown in Fig. 3 from which it may be seen that the effects of
274
275
0 Normal,
Fasted,
=Fasted,
AFasted,
Controls
Controls
Low Loctote
High Lactate
l Controls
15
30
45
MINUTES
60
FIG.
5. Effect of (+)-decanoplcarnitine
on ketogenesis and
gluconeogenesis in perfused livers from fasted rats. Livers from
24.hour fasted rats were perfused with oleic acid in the presence
or absence of 0.5 mM (t^)-decanoylcarnitine
as described under
Experimental
Procedure.
The aerfusion fluid contained nondialyzed erythrocytes.
Results are given as mean values &
S.E.M. for eight control livers and four livers with (+)-decanoylcarnitine.
MINUTES
276
In agreement with our earlier report (5) the dramatic stimulatory effect of starvation on total ketone body synthesis was
a.ccornpanied by a striking increase in the flow of label through
the ketogenic pathway and a marked decrease in its incorporation
into lipids (Lines 1 and 2, Table II).
Removal of lactate from
the medium by dialysis greatly stimulated
the initial rate of
ketogenesis (0 to 30 min) in livers from fasted rats while comparable rates were observed with both dialyzed and nondialyzed
cells over the 30- to 60.min interval (Fig. 4C). For this reason
it was to be expected that in the latter time period the distribution of oleate between oxidative and esterification
pathways
would be similar under both circumstances and, as shown in
Lines 2 and 3 of Table II, this was found to be the case. However, when high concentrations
of lactate were added its antiketogenic effect was accompanied by diversion of the fat.t,y acid
from the /3 oxidation sequence into the esterification
pat,hway
(Lines 2 and 4, Table II) in agreement with previous findings
(5). As mentioned above, removal of the lactate from the medium perfusing livers from normal animals had no effect on the
rate of ketogenesis at any time point in the perfusions.
Consistent with this finding were the essentially identical distribution
patterns of labeled oleate when these livers were perfused with
dialyzed or nondialyaed cells (Lines 1 and 5, Table II).
The most significant aspect of this series of experiments was
the finding that (+)-decanoylcarnitine,
which virtually
completely blocked fatty acid oxidation, had the effect of switching
acutely t,he mode of disposal of oleate in livers from fasted rats
to a pattern characteristic of livers from normal animals, i.e. in
the presence of the inhibitor
ketogenesis in livers from fasted
rats was reduced by 907, and virtually all of the fatty acid taken
up by the tissue was esterified (Lines 1,2, and 6, Table II).
This
observation
clearly demonstrates
that no fundamental
defect
exists in the esterifying capacity of the liver in the fasted state
and supports the view that the diminished ketogenesis and enhanced esterification
of oleate in livers from well fed animals
results primarily from a greatly depressed capacity to oxidize
long chain fatty acids. Moreover, this effect of (+)-decanoylcarnitine on the metabolism of oleic acid was exerted irrespective
of the lactate concentration in the perfusion fluid (Lines 6 and 7,
Table 11). This observation suggests that, under circumstances
where the ability to oxidize the fatty a.cid is blocked sufficient
TABLE
Metabolism
II
of [f-W]oleic acid in perfused livers from normal and E-hour fasted rats
fasted rats were perfused as at an initial
concentration
of 0.5 mM.
L-($)-Lactate
Livers
from normal and 24-hour
described
in the legends
to Figs. 3 apd 4. The concentration
of
fatty
acid in the cell-free
perfusate
was maintained
at approximately
0.75 mM by the constant
infusion
of albumin-bound
oleate.
the
30-min
State of animal
time
point.
Other
additions
pm&s/f00
1. Normal
6. Fasted
(4). .
Nondialyzed
Nondialyzed
Dialyzed
Nondialyzed
Dialyzed
Nondialyzed
7. Fasted
(2).
Dialyzed
2.
3.
4.
5.
(4).
Fasted (8).
Fasted (4).
Fasted (4).
Normal (3).
initially
to bring
its
infused
to maintain
given for the 30- to
percentage
of recovery
ber of livers shown
Lactate
(+)-Decanoylcarnitine
(+)-Decanoylcarnitine
15
89
85
53
18
8.5
12
f
f
f
f
f
f
Used
Kethes
2.0
3.6
7.5
5.0
3.2
0.9
was added
concentration
to 10 mM and was constantly
a concentration
above 7.5 mM.
Values
are
60-min
time interval
and are presented
as
of isotope
(means f S.E.M.)
for the numin parentheses.
Liver
70 [I-CJoleale/fOO
35.3
39.8
42.5
39.6
38.8
35.7
39.0
f
f
f
f
f
zk
2.6
1.1
0.5
3.1
1.8
1.1
1.4
15.6
18.7
7.5
2.3
0.9
/
1.0
f
f
f
f
f
f
0.3
1.1
4.3
0.7
0.5
0.06
YLipids
31.4
14.2
16.1
23.6
33.6
27.9
43.7
i
f
f
i
f
f
1.4
1.0
0.9
2.7
0.9
2.7
97.5
89.5
92.4
91.5
96.9
92.5
105.5
f
f
i
f
f
f
3.1
1.7
3.7
1.2
0.6
1.8
might be expected to have noticeable effects on the rate of gluconeogenesis. Thus, at the lactate concentrations
used here,
which are within the physiological range, the virtually complete
shut down in fatty acid oxidation has had little, if any, effect on
the ability of the liver to convert this substrate into glucose.
These findings appeared to contrast with those reported by
Williamson et al. (17) who showed that inhibition
of fatty acid
oxidation by ( +)-decanoylcarnitine
drastically
curtailed
the
rate of gluconeogenesis from lactate in the perfused rat liver.
Since these workers used much higher concentrations
of the
gluconeogenic substrate (8 to 12 mM) it was felt worthwhile
to
repeat our studies under conditions where the initial lactate
concentration was set at 10 InM and maintained at a high level by
constant infusion throughout
the perfusion period.
In experiments not shown ketogenesis was again virtually abolished by
the inhibitor, and over the 15- to 60-min interval glucose output
was reduced from 168 f 4.7 to 69 i 9.5 ~moles/lOO g of body
weight (means + S.E.M. of four perfusions in each group),
confirming the findings of Williamson el al. (17). It is apparent,
therefore, that fatty acid osidation is required for high but not
low rates of gluconeogenesis from lactate.
Metabolism of [I-14C]Oleic Acid in Perfused Livers from Normal
and %$-Hour Fasted Rats-As
mentioned under Experimental
Procedure, labeled oleate was added to the perfusion fluid at
the 30-min time point in the studies shown in Figs. 4 and 5. At
60 min the distribution
of the label between ketone bodies and
liver lipids was determined.
From the data compiled in Table
II, it is apparent that the quantity of isotope taken up by the
livers was approximately
the same in all experimental
groups
and was generally in the region of 40% of the amount added per
100 g of body weight.
This finding is consistent with other
reports on fatty acid metabolism in the perfused rat liver (5, 15,
19) and was also to be expected on the basis of the data shown
in Table I.
As noted previously (5), under the experimental
conditions
used here virtually all of the de novo-synthesized lipids (triglycerides and phospholipids)
were retained in the liver over the
time period studied.
The sum of the isotope recovered in lipids
and ketone bodies invariably accounted for the bulk of the fatty
By comparison, the amount of labeled oleate
a.cid metabolized.
converted into 14C02 was small and was not routinely measured.
277
ol-glycerophosphate
is supplied for its esterification
enous sources in the absence of an added precursor
phosphate.
DISCUSSION
27s
observed herr, the possibility of metabolic regulation at this site
deserves further exploration
since measurements of enzyme activity in brokeu cell preparations
may not always reflect differences prevailing iu the intact organ (27-29).
Finally, if hepatic fatty acid metabolism is controlled primarily
by activation-inact,iration
of the acylcarnitinetransferase
reaction
it is intriguing
to ask how this is related to the simultaneous
changes seen here in other physioIogica1 parameters occurring
with fast)ing and refeeding.
Since lactate exerts its antiketogenic
effect by causing a diversion of long chain fatty acids from an
osidat.ive fate int.0 the t,riglyceride-synthesizing
pathway in the
absence of significant
changes in ol-glycerophosphate
levels,
further investigation of the mechanism of action of lactate and
other antiketogenica agents (5) may provide a clue to the means
whereby dietary and hormonal control of hepatic fatty acid
metabolism and ketogenesis is accomplished.
REFERENCES
1. HE&IS, I<., Ross, B. D., BIGRRY, M. N., BND KREHS, H. A.
(1966) I?kh.e~~~. J. 101, 284
2. EXTON, J. H., CO~BIN, J. G., AND PAINT, C. R. (1969) J. LGol.
Chem.
244,
4095
3. WILLIAMSON,
J. R., BROWNING,
J. Hiol. Ch,em. 244, 4607
4. MENAHAN,
L. A., AND WINAND,
E. T.,
0.
AND SCHOLZ,
(1969) Eur.
R. (19G9)
J. Biochem.
9,
182
5.
MCGARRY,
J.
D.,
AND
FOSTER,
D. W. (1971) J. Biol.
Chem.
6247
6. May~s,
246,
8.
9.
10.
MEIER,
J.
H., .4ND
MCGARRY,
WILLIAMSON,
KREBS,
CHERNIX,
J. D., FALOONA,
G. R., UNGEI~,
R.
W: (1972) J. Lipid Res. 13; 228
J. D., AND FOSTER, Il. W. (1972) Metabolism
21,471
D. H., VICLOSO, D., ELLINGTON,
E. V., L4~~
H. A. (1969) Biochem.
J. 114, 575
S. S., AND Scow, 1-L 0. (1959)
Amer. J. Physiol.
196,
M., MCGARRY,
FOSTER,
D.
125
11.
GOOD,
Chem.
EXTON,
J. Biol.
Chem.
J. H.,
2622
15. VAN HARKEN,
J. Biol.
16.
17.
245,
.~ND
11.
Chem.
R.,
4382
PARK,
DIXON,
(1967) J. Biol.
C. W.,
AND
HEIMB~:RG,
Chem.
242,
M. (1969)
244, 2278
H. A., WALLACE,
P. G., HEMS, R., AND
R. A. (1969) Riochem.
J. 112, 595
WILLIAMSON,
J. R., BROWNING,
E. T., SCHOLZ,
R.,
R. A., lick
FRITZ,
I. B. (1968) Diabetes 17. 194
FRITZ,
i. B. (19Gl) khysiol:
Rev: 41, 52
MORRIS,
B. (1963) J. Physiol.
168, 5G4
FOSTER,
D. W. (1967) J. C&in. Invest. 46, 1283
KREUS,
18.
19.
20.
21. SBLING,
H. D., WILLMS,
B.,
J. (1968) Eur. J. Biochem.
22.
C. R.
FRIEDRICHS,
4, 364
D.,
FK.EEDL.AND,
KREISI~ERG,
AND KLICINIXE,
J. R., BROWNING,
E. T., THURMAN,
R. G., AND
SCHOLZ, K. (1969) J. Biol. Chem. 244, 5055
23. HEIMBERG,
M., V.~N HaRKEN,
D. R., AND BROWN,
T. 0. (1967)
Biochim.
Biophys.
Acta 137, 435
24. FRITZ, I. B., BND Ma~auIs,
N. R. (1965) Proc. iVat. Acud. Sci.
U. 23. A. 54, 1228
25. AUGENFELD,
J., AND FRITZ, I. B. (1970) Can. J. Biochem.
48,
288
K. R. (1965) Biochim.
Biophys.
Acta 98, 652
26. NORUM,
27. WILLuvsoN,
I>. H., BATES, M. W., ANII KIUG~S, H. A. (1968)
Biochem.
J. 108, 353
28. KREBS,
H. A. (1967) 3. Konferenz
Gesellschaft
Biologische
Chemie, p. 129, Springer-Verlag,
Berlin, Heidelberg,
and
New York
J. I)., AND FOSTXR,
D. W. (1969) Biochim.
Biophys.
29. MCGARRY,
Acta 177, 35
WILLIAMSON,
Acknow2edgnlclzts-\e
acknowledge with pleasure the expert
technical assistance of Ms. Petra Contreras and Ms. Martha
Cook. We are also iudebted to Dr. I. 13. Fritz and to Dr. Y.
Kawashima
(Otsuka Pharmaceutical
Factory,
Naruto,
ToInsulin
kushima, Japan) for gifts of ( +)-decanoylcarnitine.
assays were generously performed by the laboratory of Dr. Roger
H. Unger.
7.