Download Impaired Tricarboxylic Acid Cycle Activity in Mouse Livers Lacking

THE JOURNAL
OF
BIOLOGICAL CHEMISTRY
Vol. 279, No. 47, Issue of November 19, pp. 48941–48949, 2004
Printed in U.S.A.
Impaired Tricarboxylic Acid Cycle Activity in Mouse Livers Lacking
Cytosolic Phosphoenolpyruvate Carboxykinase*
Received for publication, June 24, 2004, and in revised form, August 30, 2004
Published, JBC Papers in Press, September 3, 2004, DOI 10.1074/jbc.M407120200
Shawn C. Burgess‡§, Natasha Hausler‡, Matthew Merritt‡, F. Mark H. Jeffrey‡, Charles Storey‡,
Angela Milde‡, Seena Koshy¶, Jill Lindner储, Mark A. Magnuson储, Craig R. Malloy‡**,
and A. Dean Sherry‡¶
From ‡The Mary Nell and Ralph B. Rogers Magnetic Resonance Center, Department of Radiology,
University of Texas Southwestern Medical Center, Dallas, Texas 75235-9085, the ¶Department of Chemistry,
University of Texas, Dallas, Texas 75083-0688, the 储Department of Molecular Physiology, Vanderbilt University
School of Medicine, Nashville, Tennessee 37232-0615, and the **Department of Internal Medicine, Veterans Affairs
North Texas Health Care System, University of Texas Southwestern Medical Center, Dallas, Texas 75235-9085
Liver-specific phosphoenolpyruvate carboxykinase
(PEPCK) null mice, when fasted, maintain normal whole
body glucose kinetics but develop dramatic hepatic steatosis. To identify the abnormalities of hepatic energy
generation that lead to steatosis during fasting, we studied metabolic fluxes in livers lacking hepatic cytosolic
PEPCK by NMR using 2H and 13C tracers. After a 4-h
fast, glucose production from glycogenolysis and conversion of glycerol to glucose remains normal, whereas
gluconeogenesis from tricarboxylic acid (TCA) cycle intermediates was nearly absent. Upon an extended 24-h
fast, livers that lack PEPCK exhibit both 2-fold lower
glucose production and oxygen consumption, compared
with the controls, with all glucose production being derived only from glycerol. The mitochondrial reductionoxidation (red-ox) state, as indicated by the NADH/
NADⴙ ratio, is 5-fold higher, and hepatic TCA cycle
intermediate concentrations are dramatically increased
in the PEPCK null livers. Consistent with this, flux
through the TCA cycle and pyruvate cycling pathways is
10- and 40-fold lower, respectively. Disruption of hepatic
cataplerosis due to loss of PEPCK leads to the accumulation of TCA cycle intermediates and a nearly complete
blockage of gluconeogenesis from amino acids and lactate (an energy demanding process) but intact gluconeogenesis from glycerol (which contributes to net NADH
production). Inhibition of the TCA cycle and fatty acid
oxidation due to increased TCA cycle intermediate concentrations and reduced mitochondrial red-ox state
lead to the development of steatosis.
Hepatic phosphoenolpyruvate carboxykinase (PEPCK)1 is a
major control point for gluconeogenesis (1). Excess PEPCK
* This work was supported in part by National Institutes of Health
Grants RR02584, U24-DK59632, and HL-34557 and a grant from the
American Diabetes Association (to M. A. M.). The costs of publication of
this article were defrayed in part by the payment of page charges. This
article must therefore be hereby marked “advertisement” in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
§ To whom correspondence should be addressed: the University of
Texas Southwestern Medical Center, Mary Nell and Ralph B. Rogers
Magnetic Resonance Center, 5801 Forest Park Rd., Dallas, TX 752359085. Tel.: 214-648-5893; Fax: 214-648-5881; E-mail: shawn.burgess@
utsouthwestern.edu.
1
The abbreviations used are: PEPCK, phosphoenolpyruvate carboxykinase; TCA, tricarboxylic acid; MAG, monoacetone glucose; PCA,
perchloric acid; ACAC, acetoacetate; BHB, ␤-hydroxybutyrate; OAA,
oxaloacetate; PEP, phosphoenolpyruvate; red-ox, reduction-oxidation
state.
This paper is available on line at http://www.jbc.org
expression in mice causes hyperglycemia (2), hyperinsulinemia, and increased glucose turnover (3). Inhibition of PEPCK
by pharmaceutical interventions causes hypoglycemia (4) and,
as expected, the global ablation of the cytosolic isoform of
PEPCK in mice by genetic manipulation results in nonviable
offspring (5). Most surprisingly, the liver-specific deletion of
cytosolic PEPCK yielded a phenotype that, except during fasting and exercise, was virtually indistinguishable from control
mice (5, 6). Even after a 24-h fast, when liver glycogen is
depleted and flux through liver PEPCK should be essential to
maintain plasma glucose, these animals are euglycemic and
glucose turnover is normal. By using NMR spectroscopy and
stable isotope tracers, we demonstrated that approximately
⬃60% of whole body glucose production in liver-specific PEPCK
knock-out animals is derived from lactate and alanine (6). This
suggests that either an alternative route to glucose production
that bypasses PEPCK in these livers exists or that the majority
of whole body gluconeogenesis is extrahepatic (6). In marked
contrast to the minimal impact that the absence of hepatic
PEPCK has on systemic glucose kinetics, these mice develop
dramatic hepatic steatosis after fasting (6) even though enzymes of the TCA cycle and ␤-oxidation are up-regulated (5) in
liver tissue. This observation suggests that cataplerosis from
the TCA cycle is tightly linked to oxidation of substrate in the
TCA cycle (7).
The purpose of the present work was to clarify the means by
which PEPCK null livers contribute to glucose production and
to determine how diminished PEPCK activity affects hepatic
energy homeostasis. Although the lipid accumulation must be
due to an imbalance of fatty acid uptake, ␤-oxidation (8, 9),
lipogenesis (10), or lipoprotein export (11), we sought to determine exactly how these processes are disturbed. Ideally, such
investigations would be done in vivo, but such studies are
complicated by the fact that other organs are known to contribute to gluconeogenesis in these animals (6). Thus, an isolated
perfused liver preparation was chosen for these experiments to
remove the influence of other organs and neurohumoral signals
from liver function. As previous results (5) indicated a dramatic
difference in metabolism depending on nutritional state, we
chose to compare metabolism in livers from mice after a 24-h
fast versus a short term 4-h fast.
Here we used NMR and a combination of 2H and 13C tracers,
isotopomer analysis, and metabolite assays to investigate the
various pathways summarized in Fig. 1. The data provide
evidence for near complete impairment of gluconeogenesis
from TCA cycle cataplerosis, a reduced mitochondrial red-ox
state, build-up of certain TCA cycle intermediates, and a
48941
48942
NMR Analysis of Livers Lacking PEPCK
FIG. 1. Hepatic glucose metabolism.
Pathways in boldface are interrogated by
combined deuterium and 13C NMR isotopomer analysis. The deuterated water
method introduces deuterium label into all
the positions of glucose, but enrichment at
H2, H5, and H6S occur at phosphoglucose
isomerase, triose-phosphate isomerase,
and fumarase making those positions diagnostic of glycogenolysis, GNGglycerol, and
GNGPEP. [U-13C3]Propionate is used as vehicle to introduce 13C label into the TCA
cycle intermediate pool by its conversion to
succinyl-CoA. The resulting 13C isotopomers that form in either glucose or tissue
glutamate are used to determine anaplerosis (assumed equal to cataplerosis,
PEPCK), pyruvate cycling, and GNGPEP
relative to TCA cycle turnover (citrate synthase). Pyruvate carboxylase and PEPCK
are presumed to be the main sites of
anaplerosis and cataplerosis, respectively.
Note that the pyruvate cycling pathway via
pyruvate kinase activity cannot be distinguished from pyruvate cycling due to malic
enzyme activity. Substrates in boxes were
supplied in the perfusate to the isolated
perfused livers.
dramatic inhibition of TCA cycle activity in the liver of PEPCK
null mice.
MATERIALS AND METHODS
Chemicals—[U-13C3]Propionate (99%) and 2H2O (99%) were purchased from Cambridge Isotopes (Andover, MA). DSC-18 solid phase
extraction gel and other common chemicals were purchased from Sigma
unless otherwise noted.
Animals—Liver-specific PEPCK knock-out mice, pcklox/lox/AlbCre
(PEPCK null), and littermate controls, pcklox/lox (control), were generated as described previously (5). 3–5-Month-old mice weighing 25–30 g
were maintained on standard laboratory chow. Prior to removal of
livers for perfusion, animals were either fasted for 24 h (long term or
LTfasted) or fasted for 12 h, fed a liquid diet of EnsureTM for 2 h, followed
by a 4-h fast (short term or STfasted). Mice were fed in this manner to
synchronize nutritional status by ensuring a 2-h feeding period, followed by a standard fast. The liquid diet also minimizes residual stomach contents.2
Liver Perfusion Experiments—All protocols were approved by the
Institutional Animal Care and Use Committee. Animals were anesthetized by intramuscular injection of ketamine (Ketaset, Aveco, Fort
Dodge, IA), and the livers were isolated and perfused as described
previously (12), but with some modifications. Briefly, a mid-line laparotomy was performed to expose the liver and portal circulatory system.
The portal vein was cannulated, and heparin (50 IU) was injected into
the portal vein to prevent the formation of blood clots. The hepatic vein
and inferior vena cava were dissected, and the perfusate flow through
the portal vein was started simultaneously with a peristaltic pump at 8
ml/min. The liver was carefully removed from the carcass and suspended in a container of effluent perfusate, which was not re-circulated
but constantly siphoned off and stored on ice for the duration of the
experiment. The perfusate was composed of Krebs-Henseleit bicarbonate buffer containing 1.5 mM lactate, 0.15 mM pyruvate, 0.25 mM glycerol, 0.2 mM octanoate, 0.5 mM [U-13C3]propionate, and 3% v/v D2O. The
livers were perfused for 60 min providing a total of 480 ml of perfusate.
For a few livers, perfusions were continued for 2 h to confirm that
isotopic steady state had been reached. After perfusion, livers were
freeze-clamped using aluminum tongs cooled to ⫺80 °C and stored in a
⫺80 °C freezer until use.
Fractions of perfusate were collected every 15 min for the assay of
glucose production and oxygen consumption. pO2 was determined on
afferent and efferent perfusate by using a blood gas analyzer (Instrumentation Laboratory, Lexington, MA), and oxygen consumption was
calculated as detailed previously (13). Glucose was assayed by enzymecoupled reactions (14).
Sample Preparation—After completion of the experiment, the perfusate was frozen until it was worked up completely. The perfusate was
2
C. Storey, A. Milde, and S. C. Burgess, unpublished observations.
evaporated under vacuum at 50 °C. The resulting material was resuspended in 10 ml of 90:10 methanol/water and stirred for 10 min. The
solution was decanted, and the step was repeated three times with all
fractions combined and dried by evaporation under vacuum. The resulting partially de-salted material was dissolved in a minimum volume of
water (about 5 ml) and passed through a column containing 20 ml of
Amberlite IRA-67 anion exchange and 20 ml of Dowex 50W-X8-200
cation exchange resins (prepared as described in the product literature)
in series. The glucose was removed with 60 ml of water; the pH was
adjusted to 6 –7, and the sample was lyophilized and stored until further workup. Amino acids were isolated from some perfusate samples
(15) to investigate alternative carbon disposal by the livers.
NMR Analysis—Purified glucose was converted to the 1,2-isopropylidene glucofuranose derivative (monoacetone glucose (MAG)) as described previously (16, 17). The MAG samples were dissolved in 160 ␮l
of high pressure liquid chromatography grade acetonitrile with 5–10 ␮l
of water and transferred to a 3-mm NMR tube. Deuterium NMR spectra
of MAG were collected using a 14.1-tesla Varian INOVA spectrometer
and 3-mm broadband probe, tuned to 92-MHz as described previously
(17, 18). A 90° pulse was applied, and the signal was acquired over 1 s
(sweep width ⫽ 1000 Hz) with no further delay. 2H NMR spectra were
typically signal-averaged for 4 h at 50 °C. 40 ␮l of deuterated acetonitrile was added to the NMR sample for locking purposes, and 13C NMR
spectra of MAG were collected as described previously (17) on the same
spectrometer and probe tuned to 150-MHz. A 50° pulse, acquisition time
of 1.5 s, and no further delay times were found to give the highest
sensitivity for the MAG carbons (19). 13C NMR spectra of PCA extracts
of the liver were recorded using a 45° pulse, 1.5-s acquisition time, and
1.5-s delay. Broad band 1H decoupling was accomplished using the
standard WALTZ-16 technique. 2H and 13C spectra were analyzed
using the curve-fitting routine supplied with NUTS PC-based NMR
spectral analysis program (Acorn NMR Inc., Freemont, CA).
Metabolic Profile—During the deuterated water experiment, the hydrogens attached to the carbon backbone of newly released glucose
become enriched with 2H at specific locations that depend upon the
synthetic pathway (Fig. 1). 2H enrichment of MAG positions H2, H5,
and H6S (as evaluated by 2H NMR) were used to determine the origin
of glucose. The 1-H2/H5 ratio, has been shown to represent the fraction
of glucose originating from glycogenolysis (20, 21). Additionally, enrichment at the H6S (H6S/H2) position reflects gluconeogenesis from carbon
units originating from the TCA cycle (oxaloacetate pool) (22, 23). The
difference in the H5 and H6S enrichment ([H5-H6]/H2) reflects gluconeogenesis from substrates feeding the gluconeogenic pathway above the
TCA cycle (presumably glycerol) (17, 19).
Pathways intersecting the TCA cycle were evaluated by 13C isotopomer analysis of glucose (i.e. MAG) or glutamate (24). The 13C NMR
multiplets in glucose or glutamate generated by the tracer
[U-13C3]propionate were evaluated to determine flux (relative to TCA
cycle flux) through anaplerosis/cataplerosis and pyruvate cycling (Fig.
1). Together the 2H and 13C data were integrated with total glucose
NMR Analysis of Livers Lacking PEPCK
48943
FIG. 2. Oxygen consumption in isolated perfused null and control livers. Oxygen consumption was higher after a long term fast (24 h) than after a
short term fast (4 h) in both the control
and PEPCK null livers. PEPCK null livers had lower oxygen consumption than
the control livers under either nutritional
state, indicating a lower energy demand.
Data are reported as the mean ⫾ S.D. for
n ⫽ 4 –5 livers.
production to determine TCA cycle flux (i.e. citrate synthase activity) as
described previously (17, 19).
Liver Tissue Analysis—Frozen liver samples were pulverized to a
fine powder in a mortar and pestle cooled with liquid N2. The powder
was agitated in 5 ml of ice cold 3% PCA in a 20-ml centrifuge tube. The
suspension was centrifuged at 13,000 ⫻ g for 10 min at 4 °C. The liquid
was decanted, and the plug was rinsed with a small volume of PCA that
was combined with the decanted portion. The extract was neutralized to
pH 7 with KOH, centrifuged at 13,000 ⫻ g for 10 min, decanted, and
then lyophilized.
The extract was dissolved in 600 ␮l of 2H2O for 13C NMR isotopomer
analysis of liver glutamate (24). The spectra of the control and null
livers were also used to qualitatively observe the pool sizes of various
metabolites that become 13C-enriched as a consequence of
[U-13C3]propionate metabolism.
About 100 mg of the liver tissue was saved for determination of
mitochondrial red-ox state by analysis of the red-ox pair acetoacetate
(ACAC) and ␤-hydroxybutyrate (BHB) (25). The tissue was extracted,
as above, with 600 ␮l of 3% PCA. After centrifugation, the extract was
neutralized with 80 ␮l of concentrated KH2PO4. The samples were kept
on ice, and ACAC and BHB were immediately evaluated by 1H NMR
(26). 400 ␮l of the extract was transferred to a 5-mm NMR tube, and an
additional 100 ␮l of a 0.5 mM sodium 2,2-dimethyl-2-silapentane-5sulfonate standard in 99% 2H2O was added. 1024 transients were
collected with a 90° pulse width and 3-s acquisition time. Concentrations were determined with external standards of BHB referenced to
the same internal 2,2-dimethyl-2-silapentane-5-sulfonate standard.
With the sodium 2,2-dimethyl-2-silapentane-5-sulfonate resonance set
to 0 ppm, the methyl resonances of BHB and ACAC had chemical shifts
of 1.20 and 2.27 ppm, respectively. The limit of detection for this
technique was about 4 ␮M. The NAD⫹/NADH ratio was determined by
assuming an equilibrium constant of 4.93 ⫻ 10⫺2 for the ␤-hydroxybutyrate dehydrogenase reaction (25).
Statistical Methods—All data are reported as the average ⫾ S.D.
Statistical analysis of the data was performed using the two-tailed
Student’s t test, assuming unequal variances. p values of less than 0.05
were considered significant.
RESULTS
A major advantage of using an isolated perfused mouse liver
preparation is that tissue oxygen consumption and glucose
production can be directly monitored as often as necessary. In
experiments described below, oxygen consumption and glucose
output were monitored every 15 min. The first 15-min perfusion period yielded sporadic values for oxygen consumption and
glucose output, but the last 3 periods provided consistent results. Thus, glucose production and oxygen consumption from
the last 3 periods (totaling 45 min) were averaged for each
liver. Oxygen consumption (Fig. 2) was lower in livers from
PEPCK null mice compared with littermate controls under
both STfasted (controlled short term fast, see “Materials and
Methods”) and LTfasted (long term fast, see “Materials and
Methods”) conditions (p ⬍ 0.02, p ⬍ 0.03 respectively). Oxygen
consumption was also lower in STfasted versus LTfasted livers
from both null and control animals.
Glucose Production—Glucose production was twice as high
in control livers versus PEPCK null livers in LTfasted conditions
and also tended to be higher in the STfasted condition (Fig. 3). In
general, livers from STfasted mice had higher glucose output
than livers from LTfasted mice regardless of genotype. Glucose
production may have contributions from glycogen stores (i.e.
glycogenolysis) or from gluconeogenesis. The carbon supply for
gluconeogenesis can be provided from substrates such as glycerol that feed directly into the triose pool (GNGglycerol) or from
substrates such as lactate, pyruvate, or amino acids that must
undergo anaplerosis via pyruvate carboxylase to oxaloacetate
(OAA) followed by cataplerosis via PEPCK to phosphoenolpyruvate (PEP) as the first steps of glucose synthesis (GNGPEP).
The sources of glucose in this study were elucidated by using
the deuterated water method (17, 18, 20, 23, 27). This technique most often uses mass spectrometry as the analytical tool
but has been shown to work equally well when monitored by 2H
NMR (18). Fig. 4 shows examples of 2H NMR spectra of MAG
derived from glucose generated by livers of control and PEPCK
null mice. The spectrum of MAG from the PEPCK null liver is
clearly lacking substantial 2H in the H6 positions compared
with the control. Because chemical exchange at the H6R and
H6S occur at the level of alanine aminotransferase and fumarase, respectively (23), the low enrichment at these sites reflect
severely limited gluconeogenesis supplied by cataplerotic flux
from the TCA cycle. This is in agreement with the expected
phenotype of a liver lacking PEPCK. These 2H NMR data were
used to determine the contribution of PEP (formed from lactate
or amino acids) to glucose production (i.e. GNGPEP). GNGPEP
was lower by 11-fold in PEPCK null livers compared with
controls under STfasted conditions (Fig. 3A) and 18-fold under
LTfasted (Fig. 3B) conditions.
Although gluconeogenesis due to cataplerosis from the TCA
cycle is insignificant in PEPCK null livers, gluconeogenesis
from glycerol (GNGglycerol) was maintained. Under STfasted conditions, GNGglycerol was significantly higher in control livers
(0.61 ␮mol/min/g) than in null livers (0.31 ␮mol/min/g) (Fig.
3A). There was no significant difference in GNGglycerol under
LTfasted conditions between control livers (0.54 ␮mol/min/g)
48944
NMR Analysis of Livers Lacking PEPCK
FIG. 3. Glucose metabolism in the isolated perfused livers of
short term fasted (4 h, STfasted) (A) and long term fasted (B) (24
h, LTfasted) mice. Hepatic sources of glucose production (GP) were as
determined by 2H NMR (as in Fig. 4) and glucose assay of effluent
perfusate. Glycogenolysis (Gly) represents flux from glycogen to glucose. GNGglycerol represents flux from glycerol to glucose. GNGPEP represents flux from PEP (i.e. cataplerosis of TCA cycle intermediates via
PEPCK) to glucose. As expected, GNGPEP is almost completely suppressed in the PEPCK null livers under both nutritional states.
GNGglycerol did not supplement gluconeogenic capacity in the PEPCK
null liver, and total glucose production was lower in the PEPCK null
liver after the long term fast. Data are reported as the mean ⫾ S.D. for
n ⫽ 4 –5 livers.
and PEPCK null livers (0.37 ␮mol/min/g) (Fig. 3B). Whereas
GNGglycerol is less or equal in PEPCK null livers compared with
controls, this source represented 92% of the glucose produced
in the null liver compared with only 53% in the control livers.
This result is in agreement with our earlier observation that
the PEPCK null liver maintains its capacity to make glucose
from glycerol (6).
Glycogenolysis was low in livers of both PEPCK null and
control mice in the LTfasted state (Fig. 3B). In contrast, PEPCK
null and control livers showed substantial and equal rates of
glycogenolysis, 0.69 and 0.73 ␮mol/min/g, respectively, under
STfasted conditions (Fig. 3A). Whereas absolute glycogenolysis
was about the same in null and control livers, glycogen contributed a larger percentage of the glucose produced by null livers
(68%) than control livers (49%) in the STfasted state.
TCA Cycle Associated Pathways—Gluconeogenesis is intimately linked to the TCA cycle by both energy demand and
substrate supply. To evaluate potential differences in TCA
cycle fluxes, [U-13C3]propionate was used as a vehicle to deliver
a 13C tracer into the TCA cycle via succinyl-CoA (28). The
result is formation of multiple positional isomers of labeled
intermediates commonly referred to as 13C isotopomers. The
population of groups of 13C isotopomers is reflected by multi-
plets due to spin-spin coupling between adjacent 13C nuclei (24)
in all metabolites associated with the TCA cycle, and the relative areas of these multiplets is determined by the activity of
pathways that intersect in the TCA cycle (e.g. citrate synthase,
anaplerosis/cataplerosis, pyruvate cycling, and gluconeogenesis) (24). The combined information reported by the 2H and 13C
NMR spectra plus absolute glucose production rates may then
be used to determine absolute flux through PEPCK, pyruvate cycling (OAA 3 PEP 3 pyruvate 3 OAA or malate 3
pyruvate 3 OAA) and the TCA cycle (17, 19).
Fluxes Relative to TCA Cycle Turnover—13C NMR spectra of
MAG from effluent glucose synthesized by control and PEPCK
null livers are shown in Fig. 5A. The C2 resonance of MAG from
control livers showed multiplets typical of cataplerosis of TCA
cycle intermediates into glucose. In contrast, the 13C spectrum
of MAG derived from PEPCK null livers was almost devoid of
multiplets, consistent with a lack of gluconeogenesis from TCA
cycle intermediates. The low levels of multiplets in the C2
resonance of MAG from PEPCK null livers precluded a determination of anaplerotic/cataplerotic, gluconeogenic, and pyruvate cycling fluxes in the usual way (24). However, given that
glucose and glutamate share the common TCA cycle intermediate, OAA, we turned our attention to analysis of the multiplets in the C2 resonance of liver glutamate (17, 24). This
requires the assumption that a loss of cytosolic PEPCK activity
has no impact on exchange of carbons between oxaloacetate
and glutamate. The C2 resonances from the 13C NMR spectra
of glutamate isolated from control and PEPCK null livers are
also given in Fig. 5A. As anticipated, the C2 resonances of
glutamate and glucose from control livers were similar in appearance, and an isotopomer analysis of each resonance gave
identical measures of anaplerosis, gluconeogenesis, and pyruvate cycling flux (Fig. 6). This same analysis when applied to
the C2 resonance of glutamate from PEPCK null livers indicated that anaplerosis and cataplerosis (relative to TCA cycle
flux) were reduced by 3-fold and that pyruvate cycling flux
(relative to TCA cycle flux) was suppressed ⬃10-fold in livers
from LTfast animals (Fig. 6). Pyruvate cycling flux was reduced
less (⬃4-fold) in livers after a STfast (not shown).
By definition, anaplerosis must equal cataplerosis, whereas
the difference between cataplerosis and pyruvate cycling is net
output (Fig. 1). This output is normally assumed equal to
gluconeogenesis because flux through PEPCK is considered the
overwhelming site of cataplerosis in the liver (24). However,
this is not true in PEPCK null mice, so output in this case could
reflect some intermediate associated with the TCA cycle other
than PEP. Fig. 5B shows the 13C NMR spectra of extracts of the
control and PEPCK null livers. A comparison of the two spectra
shows a dramatic increase in TCA cycle intermediate concentrations in PEPCK null livers compared with controls. The
most dramatic increases are in malate, aspartate, glutamate,
and fumarate, but citrate and succinate also appear to have
increased (these changes were not quantitated). As summarized in Fig. 6, the glutamate isotopomer analysis reported that
total cataplerosis (relative to TCA cycle flux) was lower by
⬃66% in livers from PEPCK null livers compared with controls.
The residual cataplerosis of ⬃33% may be surprising given that
GNGPEP is almost zero. The liver may also export amino acids
(29), so effluent perfusate was fractionated into amino acids,
organic acids, and neutral molecules by ion exchange chromatography to determine whether the increased tissue concentration of some intermediates might lead to increased export in
the perfusate. 13C NMR and coupled enzymatic assay analysis
of those fractions showed that 13C-labeled glutamine was exported nearly equally in PEPCK null and control LTfasted livers
(6.7 and 9.0 nmol/min/g, respectively). Other labeled metabo-
NMR Analysis of Livers Lacking PEPCK
48945
FIG. 4. Deuterium NMR spectra of MAG derived from glucose produced by LTfasted control (top) and PEPCK null (bottom) livers.
Peak areas correspond to the relative deuterium enrichment of the seven hydrogen positions of glucose. H2, H5, and H6S enrichments are
diagnostic of glycogenolysis (1-H5/H2), GNGglycerol ((H5-H6S)/H2)), and GNGPEP (1-(H5-H6S)/H2). The diagram shows how H6S enrichment occurs
during the fumarase reaction in the TCA cycle but that those intermediates cannot enter the gluconeogenic pathways in the PEPCK null livers,
resulting in low H6S enrichment in glucose produced from PEPCK null livers. Note that enrichment at the H6R position occurs during the
pyruvate-alanine equilibrium catalyzed by alanine aminotransferase and is also very low in glucose from PEPCK null livers.
lites were observed but in much lower concentrations. Thus,
the combined export of molecules other than glucose cannot
account for the cataplerotic flux ratio reported by the 13C NMR
spectrum of glutamate from PEPCK null livers, because we
would expect an efflux about 0.13 ␮mol/min/g (33% of control
GNGPEP according to Fig. 6).
Absolute Rates of Flux—Alternatively, the anaplerotic flux
ratio (anaplerosis/TCA cycle flux) as measured by NMR may
not reflect a high anaplerotic and cataplerotic flux but rather
low TCA cycle flux. This alternative explanation was tested by
converting all relative flux values into absolute flux values by
referencing them to glucose production (19). The absolute TCA
cycle flux values derived from 13C NMR, 2H NMR, and glucose
production data are compared in Fig. 7. TCA cycle flux estimated by the combined tracer experiments tended to be lower
(p ⬍ 0.07) in LTfast controls (0.10 ␮mol/min/g) versus STfast
controls (0.18 ␮mol/min/g), opposite the trend reported by oxygen consumption in these same groups (Fig. 2). Remarkably,
TCA cycle flux in the PEPCK null livers was ⬃10-fold lower
compared with controls in either nutritional state (Fig. 7).
Although oxygen consumption was lower in PEPCK null livers
in both STfast and LTfast states compared with controls, the
magnitude of the difference was not as large as that measured
for TCA cycle flux.
Of the cataplerotic flux from OAA to PEP, a portion is converted to glucose (GNGPEP), and a portion is converted to
pyruvate and then back into OAA by a process known as
pyruvate cycling (see Fig. 1). By using a similar analysis as
described above for TCA cycle flux, absolute pyruvate cycling
flux was also determined. This flux was substantial in control
livers after a STfast (0.43 ␮mol/min/g) and after a LTfast (0.20
␮mol/min/g) but was greatly reduced in PEPCK null livers in
both nutritional states (0.01 and 0.005 ␮mol/min/g, respectively). The ⬃40-fold reduced pyruvate cycling flux in PEPCK
null livers is consistent with impaired total cataplerosis (Fig. 6)
and TCA cycle activity (Fig. 7) in livers lacking PEPCK.
Mitochondrial Red-ox State—Perturbations of oxygen consumption, gluconeogenesis (30), ␤-oxidation, lipogenesis (10),
and intermediate pool sizes (31) could all be related to changes
in the cellular red-ox state of liver from PEPCK null mice.
Because the cytosolic red-ox state was presumably clamped by
the 10:1 lactate/pyruvate supplied in the perfusate in these
experiments (see “Materials and Methods”), mitochondrial
red-ox was measured using the ACAC:BHB redox pair (25).
Table I summarizes the tissue content of ACAC and BHB and
the calculated mitochondrial NADH/NAD⫹ ratio. Remarkably,
ACAC under LTfast conditions was reduced 5-fold in PEPCK
null livers compared with controls, whereas BHB was unchanged, and this translates into a 5-fold increase in the
NADH/NAD⫹ ratio in livers lacking PEPCK.
DISCUSSION
A global knock out of PEPCK in mice is lethal shortly after
birth presumably because of insufficient glucose production
from lactate/pyruvate/alanine during fasting (5). However, a
liver-specific knock out of PEPCK is not lethal; such animals
maintain euglycemia even after a 24-h fast apparently due to
glucose production from other tissues (5, 6). Despite two prior
studies that have explored the metabolic consequences resulting from the lack of PEPCK, many aspects of the phenotype of
these mice are still a mystery. Indeed, these animals provide an
experimental model that, when combined with the NMR tracer
methodology, allows the consequences of altered hepatic energy
balance to be further understood. In the current study, we
explored the metabolic basis for the marked steatosis that
occurs upon fasting in mice that lack hepatic PEPCK by using
a preparation of isolated, perfused livers. We found that energy
homeostasis is radically altered in livers lacking PEPCK as
indicated by an abnormal mitochondrial hepatic red-ox state,
increased TCA cycle intermediate pool sizes, reduced oxygen
consumption, and a dramatically lower TCA cycle flux. These
findings indicate that steatosis develops because of marked
inhibition of fatty acid oxidation and that hepatic TCA cycle
flux is an acutely sensitive interface for ␤-oxidation and
gluconeogenesis.
Hepatic Glucose Production—PEPCK null livers maintain a
relatively normal ability to produce glucose from glycogen and
glycerol but have a severely limited ability to make glucose
from TCA cycle cataplerosis. In the STfasted state, the difference
in glucose production between PEPCK null and control livers
did not differ significantly. Glycogenolysis was identical in
PEPCK null and control livers in the STfasted state even though
the glycogen content of PEPCK null livers has been reported to
be lower (6). This could reflect an enhanced glycogen phosphorylase state in the PEPCK null liver that may be necessary to
help maintain normal glucose output in the STfasted state. We
48946
NMR Analysis of Livers Lacking PEPCK
FIG. 5. Carbon-13 NMR spectra. A,
spectra of the C2 region of glucose-derived
MAG from the perfusate effluent (top
row) or liver glutamate (bottom row) from
LTfasted control (left column) and PEPCK
null (right column) livers. No multiplets
are observed in the MAG C2 region of
the PEPCK null sample because 13 C
originates in the TCA cycle (via
[U-13C3]propionate 3 succinyl-CoA). 13C
NMR spectra of glutamate from liver extracts display multiplets in the C2 region
of both the control (left column) and
PEPCK null (right column) samples. Multiplets are labeled to identify the spinspin coupling responsible for each signal.
For instance, D12 is the doublet present
when C1 and C2 (but not C3) are 13Clabeled; D23 is the doublet present when
C3 and C2 (but not C1) are 13C-labeled,
and Q is the “quartet” present when C1,
C2, and C3 are all 13C-labeled. The populations of these isotopomers are diagnostic of fluxes into and out of the TCA cycle
and should yield the same results regardless of whether glucose C2 or glutamate
C2 is analyzed because glucose carbons 1,
2, and 3 are equivalent to glutamate carbons 3, 2, and 1, respectively (assuming
the same OAA pool is responsible for
both). B, inspection of the entire 13C NMR
spectra of the liver tissue extract reveals
increased concentrations of TCA cycle intermediates in the PEPCK null livers
compared with controls.
FIG. 6. Relative flux of pathways intersecting the TCA cycle in fasted livers. Analysis of the C2 MAG or glutamate
yields total anaplerosis from all pathways
(assumed to be equal to total cataplerosis), pyruvate cycling (see Fig. 1), and output (the difference between cataplerosis
and pyruvate cycling; usually considered
gluconeogenesis). It is important to note
that 13C NMR isotopomer analysis, by itself, gives flux ratios that are relative to
TCA cycle turnover and not absolute flux.
Sustained output or gluconeogenesis relative to the TCA cycle, despite the absence of PEPCK and GNGPEP, implies
that absolute TCA cycle flux is decreased
in the PEPCK null livers. Data are reported as the mean ⫾ S.D. for n ⫽ 4 –5
livers.
NMR Analysis of Livers Lacking PEPCK
48947
FIG. 7. TCA cycle flux in control and
PEPCK null livers determined from
the combined 2H and 13C NMR data
and absolute glucose production.
PEPCK null livers had dramatically impaired TCA cycle flux under both nutritional conditions. Data are reported as
the mean ⫾ S.D. for n ⫽ 4 –5 livers.
TABLE I
Mitochondrial redox state in fasted livers from control and
PEPCK null mice
Acetoacetate and ␤-hydroxybutyrate concentrations (␮mol/g wet
weight) were determined by 1H NMR of liver tissue extracts. NAD⫹/
NADH was determined from the acetoacetate/hydroxybutyrate ratio
and an assumed equilibrium constant of 0.0493 for the BHB dehydrogenase reaction.
Control
PEPCK null
ACAC
BHB
BHB/ACAC
NAD⫹/NADH
0.19 ⫾ 0.07
0.055 ⫾ 0.005
0.19 ⫾ 0.06
0.18 ⫾ 0.08
1.0
5.1
20.1
4.0
showed previously (6) that PEPCK null livers perfused with 5
mM glycerol alone produce nearly four times as much glucose as
control livers. However, in the present study, using more physiological levels of glycerol plus other substrates, glucose production from glycerol was either identical (LTfast) or lower
(STfast) in PEPCK null livers compared with controls. This
suggests that the livers of mice lacking PEPCK do not compensate by producing more glucose from glycerol.
In the LTfasted state, glucose production is diminished by 50%
in PEPCK null livers compared with controls (p ⬍ 0.05), in
contrast to observations in vivo where glucose turnover did not
differ between PEPCK null mice and littermate controls under
LTfasted conditions (6). Glucose production in the LTfasted state
is completely supported by GNGglycerol in PEPCK null livers,
whereas the control livers have about an equal contribution
from GNGglycerol and GNGPEP. This again is in contrast to the
in vivo experiment where 60% of total body glucose was derived
from GNGPEP in the liver-specific PEPCK null mice. This,
together with the diminished glucose production in PEPCK
null livers, supports the existence of an extrahepatic source of
glucose production in the intact animal.
Glyceroneogenesis and the PEPCK Null Liver—Glyceroneogenesis is often described as an “abbreviated” form of gluconeogenesis whereby the ultimate product is 3-glycerol phosphate
rather than glucose (32). Just as in gluconeogenesis, PEPCK is
a key regulatory enzyme in glyceroneogenesis. As reviewed by
Reshef et al. (32), this process is necessary to support the
triglyceride cycle, in which plasma free fatty acids are taken up
by the liver and condensed with 3-glycerol phosphate to yield
triglycerides. The importance of this cycle and its potential
impact on hepatic cataplerosis are highlighted by reports that
75% of systemic free fatty acids are re-esterified and that the
liver may account for most of this activity (32). Moreover,
during fasting it is reported that a majority of 3-glycerol phosphate used for triglyceride synthesis in the liver is derived from
glyceroneogenesis despite the presence of both exogenous glycerol and hepatic glycerol kinase (33). Thus, it is conceivable
that a disruption in PEPCK activity might influence this cycle,
resulting in perturbed hepatic lipid metabolism and steatosis.
For instance, the role of PEPCK in glyceroneogenesis and the
triglyceride cycle has been implicated as an important factor in
the development of obesity and diabetes due to its influence on
the storage and release of fatty acids (34). The evidence suggests that, in adipocytes, increased PEPCK activity leads to
increased triglyceride storage and decreased free fatty acid
release (due to increased triglyceride re-synthesis), whereas
decreased activity leads to lipoatrophy and high plasma free
fatty acid levels (due to decreased triglyceride re-synthesis)
(34). Although glyceroneogenesis is clearly an important pathway in the liver, representing about 10% of the flux of gluconeogenesis (33), the effects of PEPCK expression on hepatic triglyceride metabolism seem less predictable than in the
adipocyte. For example, mice overexpressing PEPCK 2-fold in
the liver show little difference in plasma free fatty acid levels,
plasma triglycerides, or hepatic triglycerides compared with
littermate controls after a 6-h fast (3). Yet in the liver-specific
PEPCK null mice studied here, there are dramatic increases in
hepatic triglycerides (100%), plasma free fatty acids (60%), and
plasma triglyceride levels (34%) after a 24-h fast (5). It has
been suggested that altered hepatic lipid metabolism in the
PEPCK null mice may stem from a loss of glyceroneogenesis
and interruption of the triglyceride cycle (34, 35). Although
increased triglyceride accumulation in the face of abolished
PEPCK appears to be opposite the observations in adipocytes
(34), the number of metabolic pathways that interact with
anaplerosis and cataplerosis is much greater in the liver than
in the adipocyte. The current data offer little to support or
challenge the role of glyceroneogenesis and the triglyceride
cycle in the development of hepatic steatosis in these animals
but rather suggest that a substantial alteration in hepatic
energy homeostasis may be responsible.
Energy Homeostasis in the PEPCK Null Liver—Although the
role of PEPCK in the intermediary metabolism of gluconeogenesis is well established, the data reported here clearly support
a broader role of PEPCK in energy homeostasis via cataplerosis, which might include contributions to both gluconeogenesis
and glyceroneogenesis. Livers lacking PEPCK display an abnormal energy homeostasis as follows: 1) a modest decrease in
oxygen consumption, 2) an altered mitochondrial red-ox state,
and 3) a 10-fold decrease in TCA cycle flux. The 5-fold increase
in BHB/ACAC in PEPCK null livers reflects a 5-fold increase in
the NADH/NAD⫹ ratio in the mitochondria of these livers.
Similar changes in mitochondrial red-ox state have been implicated in other models of hepatic steatosis (36). A high level of
mitochondrial NADH is known to inhibit ␤-oxidation (8) and
TCA cycle flux (37) and stimulate triacylglycerol formation
(36). Overproduction of NADH may be related to an absence of
48948
NMR Analysis of Livers Lacking PEPCK
PEPCK in three possible ways. First, the normally very active,
NADH-consuming, glyceraldehyde-3-phosphate dehydrogenase reaction has little importance in the PEPCK null liver
because PEPCK is not actively feeding triose units into gluconeogenesis. Second, the normal demand for ATP/GTP required to convert OAA into PEP is absent, so there is less
energy demand to drive ATP synthesis in the PEPCK null liver.
Third, gluconeogenesis from glycerol, which is maintained in
the PEPCK null liver, generates excess cytosolic NADH, and
this exacerbates the reduced energy needs of the PEPCK null
liver. A direct consequence of a highly reduced mitochondrial
red-ox state is that all NADH-producing steps in the TCA cycle
are inhibited. This is reflected by the dramatically reduced
TCA cycle flux observed here by using combined glucose production rates and 2H and 13C tracers.
In addition to NADH/NAD, TCA cycle activity is also controlled by the concentration of cycle intermediates. For instance, citrate and succinyl-CoA are known to inhibit citrate
synthase and ␣-ketoglutarate dehydrogenase (37), whereas oxaloacetate allosterically inhibits isocitrate dehydrogenase. Earlier we reported (5) a 10-fold increase in hepatic malate concentrations in these mice, and during the course of the current
experiments we noted a similar increase in many other TCA
cycle intermediates as detected by 13C NMR of the liver extracts (Fig. 5B). The impact of PEPCK on the concentration of
TCA cycle intermediates is a direct reflection of its importance
as a cataplerotic pathway. Under steady-state conditions
anaplerosis and cataplerosis must always be balanced to maintain constant levels of TCA cycle intermediates (38). With the
loss of PEPCK, the most important cataplerotic pathway in the
liver, cycle intermediates accumulate because export of OAA is
blocked. The resultant increase in TCA cycle intermediate concentration causes product and allosteric inhibition of cycle activity and fatty acid oxidation.
One additional interesting observation was that the decrease
in oxygen consumption in livers lacking PEPCK does not seem
to parallel the reduction in TCA cycle activity. Oxygen consumption in the PEPCK null liver is certainly lower than
control livers in both nutritional states but not to the extent
predicted by the reduction in TCA cycle flux. However, it
should be noted that ⬃20% of hepatic oxygen consumption may
result from other oxygen-consuming reactions not linked to
oxidative phosphorylation, as evidenced by the fact that hepatic oxygen consumption decreases only ⬃80% upon the addition of potassium cyanide to a perfused rat liver (39). Thus,
one should not expect that the demand for residual oxygen
consumption in a PEPCK null liver would be necessarily less
than in control livers because a disproportionate amount of
oxygen consumption in liver is due to reactions not linked to
NADH disposal.
Oxidative phosphorylation couples the transfer of electrons
from NADH to oxygen with the associated production of ATP,
but the NADH need not originate from the reactions of the TCA
cycle. For example, ␤-oxidation coupled to the export of ketones
generates mitochondrial NADH without involving the reactions of the TCA cycle. Upon removal of all substrates from
perfusion buffer of perfused rat livers, oxygen consumption
falls only 33% (40), yet other parameters of metabolic activity
can fall by as much as 80% of basal activity (41). Here we
observe a 90 and 70% decrease in the TCA cycle activity in
STfasted and LTfasted PEPCK null livers compared with controls,
whereas oxygen consumption decreased by 50 and 20% respectively. Because ketone production in the perfused rat liver is
known to vary between 0.17 and 0.60 ␮mol/min/g (converted
from dry weight in Ref. 42), under conditions similar to those
used here, we anticipate that oxygen consumption reflects a
combination of TCA cycle activity and ketogenesis. Although
decreased oxygen consumption indicates that both lipid oxidation and TCA cycle flux is impaired in the PEPCK null livers,
TCA cycle flux appears to be exquisitely sensitive to impaired
gluconeogenesis.
These data suggest that hepatic steatosis in the PEPCK null
liver results from decreased lipid oxidation rather than increased de novo lipogenesis. This is also supported by the
observation that pyruvate cycling is greatly reduced in the
PEPCK null liver. Pyruvate cycling (Fig. 1) refers to any pathway involving combined carboxylation/decarboxylation of pyruvate. The two most common examples include pyruvate 3
OAA 3 PEP 3 pyruvate (PC ⫹ PEPCK ⫹ PK) or pyruvate 3
OAA 3 malate 3 pyruvate (PC ⫹ malate dehydrogenase ⫹
malic enzyme). The first pathway cannot exist in PEPCK null
livers, but the second pathway could be of particular importance in these livers. Lipid biosynthesis in the liver involves
transport of excess acetyl-CoA to the cytosol via citrate where it
is cleaved by ATP-citrate lyase to acetyl-CoA and OAA. AcetylCoA can then be used for lipogenesis, and OAA must take a
circuitous route back to the mitochondrial TCA cycle. OAA is
converted first into malate (via malate dehydrogenase) and
then pyruvate (via the malic enzyme), which then returns to
the mitochondria before its conversion back to OAA by
pyruvate carboxylase (37). This pathway links lipogenesis
with pyruvate cycling and cannot be distinguished from the
pyruvate cycling pathway involving PEPCK. Of the 14 NADPH
molecules required for palmatate synthesis, 8 can be generated
by the malic enzyme in converting excess malate into pyruvate
(37). The fact that we see little pyruvate cycling activity in the
PEPCK null liver suggests that lipogenesis is not an important
mechanism in producing steatosis in these livers. One cannot,
however, rule out participation of the pentose phosphate pathway in generation of NADPH for lipogenesis in these livers.
Validation of the Deuterated Water Method—The deuterated
water method for the analysis of gluconeogenesis takes advantage of hydrogen exchange at known enzymatic steps along the
pathway toward glucose production. In its simplest form, the
method uses the glucose H2/H5 deuterium enrichment ratio to
evaluate glycogenolysis versus total gluconeogenesis as sources
of glucose production (21, 23). The method has been used under
a variety conditions, including diabetes and obesity (43) in
children (44) and premature infants (45). Additionally, the 2H
NMR method offers a convenient method to differentiate further the contributions of glycerol and TCA cycle intermediates
to gluconeogenesis by detecting 2H enrichment at H6S (incorporated at the level of fumarase) versus H5 (incorporated at the
level of triose-phosphate isomerase) (17, 19). This valuable
method of quantitating glycerol contribution to gluconeogenesis has largely been overlooked, in part because its contribution
is considered small in normal humans (17) and in rats (19, 46)
but also because there has been no good way to validate the
method. We now know that the contribution of glycerol to
gluconeogenesis is actually quite high in the mouse (20 –30%)
(6) and abnormally high in lipodystrophic patients on human
immunodeficiency virus therapy (47). The PEPCK null liver
offers a unique opportunity to validate the method because
here the liver is largely incapable of generating glucose from
TCA cycle intermediates. Indeed, glucose generated by these
isolated livers has very little 2H enrichment at the H6 positions
(Fig. 4), confirming for the first time that H5 and H6 are
differentially enriched during the 2H2O experiment according
to gluconeogenic contributions from glycerol and TCA cycle
intermediates, respectively.
Technical Considerations—The methodology employed here
couples the deuterated water method with conventional 13C
NMR Analysis of Livers Lacking PEPCK
NMR isotopomer analysis. Most of the standard assumptions
have already been detailed (17, 21, 23, 24), but it is worth
mentioning a few here. First, in order to couple the TCA cycle
and pyruvate cycle activities with glucose production, it is
assumed that anaplerotic influx into the TCA cycle is equally
balanced by a cataplerotic flux. Normally in liver, this cataplerotic flux is considered equal to gluconeogenic flux, but in livers
lacking PEPCK, this might represent some other pathway. The
only other 13C-enriched metabolites identified in the perfusate
in these experiments were glutamine and lactate, but their
amount was small and no different from the control livers.
Second, the small (5%) contribution of PEP to glucose production in PEPCK null liver was estimated by measuring the areas
of small H6R peaks in the 2H NMR spectrum (Fig. 4). This 5%
estimate is consistent with the known mitochondrial PEPCK
content of mouse liver (roughly 2%). Even though the error
associated with measuring the areas of such small resonances
can be large, we estimate that it translates into an error in our
metabolic flux measurements of no more than 20%. Even with
this level of error, it is clear that TCA cycle flux and pyruvate
cycle flux are both significantly lower in livers lacking PEPCK.
Conclusions—The metabolic consequences of eliminating
PEPCK from liver are in many ways predictable, yet the impact
of this genetic manipulation on energy homeostasis in liver was
largely unanticipated. Livers lacking PEPCK continue to produce glucose from other substrates but during long term fasting cannot compensate by using glycerol as the sole gluconeogenic precursor. Removal of the energy-demanding PEPCK
pathway results in an increase in mitochondrial red-ox state
and dramatic inhibition of the TCA cycle. This results in reduced oxidation of fats via ␤-oxidation, less production of ketones, and subsequent hepatic steatosis. The isolated perfused
liver model also provided a unique opportunity to validate one
aspect of the 2H NMR method, i.e. that 2H enrichment at H6
accurately reflects gluconeogenesis from the level of the TCA
cycle. The combined 2H and 13C tracer methods verified that
livers lacking PEPCK cannot generate glucose from TCA cycle
precursors, provided an upper limit for flux from glycerol to
glucose, and provided unique insights into metabolic relationships between TCA cycle flux and glucose production in liver.
Acknowledgment—We thank Erin Smith for the excellent technical
assistance.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
REFERENCES
1. Hanson, R. W., and Reshef, L. (1997) Annu. Rev. Biochem. 66, 581– 611
2. Valera, A., Pujol, A., Pelegrin, M., and Bosch, F. (1994) Proc. Natl. Acad. Sci.
U. S. A. 91, 9151–9154
3. Sun, Y., Liu, S., Ferguson, S., Wang, L., Klepcyk, P., Yun, J. S., and Friedman,
J. E. (2002) J. Biol. Chem. 277, 23301–23307
4. DiTullio, N. W., Berkoff, C. E., Blank, B., Kostos, V., Stack, E. J., and
Saunders, H. L. (1974) Biochem. J. 138, 387–394
5. She, P., Shiota, M., Shelton, K. D., Chalkley, R., Postic, C., and Magnuson,
M. A. (2000) Mol. Cell. Biol. 20, 6508 – 6517
6. She, P., Burgess, S. C., Shiota, M., Flakoll, P., Donahue, E. P., Malloy, C. R.,
Sherry, A. D., and Magnuson, M. A. (2003) Diabetes 52, 1649 –1654
7. Magnuson, M. A., She, P., and Shiota, M. (2003) J. Biol. Chem. 278,
32485–32488
8. Bremer, J., and Wojtczak, A. B. (1972) Biochim. Biophys. Acta 280, 515–530
9. Geneve, J., Hayat-Bonan, B., Labbe, G., Degott, C., Letteron, P., Freneaux, E.,
39.
40.
41.
42.
43.
44.
45.
46.
47.
48949
Le Dinh, T., Larrey, D., and Pessayre, D. (1987) J. Pharmacol. Exp. Ther.
242, 1133–1137
Fellenius, E., Bengtsson, G., and Kiessling, K. H. (1973) Acta Chem. Scand. 27,
2893–2901
Venkatesan, S., Ward, R. J., and Peters, T. J. (1988) Biochim. Biophys. Acta
960, 61– 66
Colet, J.-M., and Muller, R. N. (1994) Magn. Reson. Mat. Phys. Biol. Med. 2,
303–305
Neely, J. R., Liebermeister, H., Battersby, E. J., and Morgan, H. E. (1967)
Am. J. Physiol. 212, 804 – 814
Kunst, A., Draeger, B., and Ziegenhorn, J. (1984) in Methods of Enzymatic
Analysis (Bergmeyer, H. U., ed) 3rd Ed., pp. 163–172, VCH Verlagsgesllschaft, Weinheim, Germany
Sherry, A. D., Zhao, P., Wiethoff, A. J., Jeffrey, F. M., and Malloy, C. R. (1998)
Am. J. Physiol. 274, H591–H599
Burgess, S. C., Weis, B., Jones, J. G., Smith, E., Merritt, M. E., Margolis, D.,
Dean Sherry, A., and Malloy, C. R. (2003) Anal. Biochem. 312, 228 –234
Jones, J. G., Solomon, M. A., Cole, S. M., Sherry, A. D., and Malloy, C. R. (2001)
Am. J. Physiol. 281, E848 –E856
Burgess, S. C., Nuss, M., Chandramouli, V., Hardin, D. S., Rice, M., Landau,
B. R., Malloy, C. R., and Sherry, A. D. (2003) Anal. Biochem. 318, 321–324
Jin, E. S., Jones, J. G., Merritt, M., Burgess, S. C., Malloy, C. R., and Sherry,
A. D. (2004) Anal. Biochem. 327, 149 –155
Schumann, W. C., Gastaldelli, A., Chandramouli, V., Previs, S. F., Pettiti, M.,
Ferrannini, E., and Landau, B. R. (2001) Anal. Biochem. 297, 195–197
Landau, B. R., Wahren, J., Chandramouli, V., Schumann, W. C., Ekberg, K.,
and Kalhan, S. C. (1996) J. Clin. Investig. 98, 378 –385
Kuwajima, M., Golden, S., Katz, J., Unger, R., Foster, D., and McGarry, J.
(1986) J. Biol. Chem. 261, 2632–2637
Landau, B. R., Wahren, J., Chandramouli, V., Schumann, W. C., Ekberg, K.,
and Kalhan, S. C. (1995) J. Clin. Investig. 95, 172–178
Jones, J. G., Naidoo, R., Sherry, A. D., Jeffrey, F. M., Cottam, G. L., and
Malloy, C. R. (1997) FEBS Lett. 412, 131–137
Williamson, D. H., Lund, P., and Krebs, H. A. (1967) Biochem. J. 103, 514 –527
Chung, Y., and Jue, T. (1992) Biochemistry 31, 11159 –11165
Wajngot, A., Chandramouli, V., Schumann, W. C., Ekberg, K., Jones, P. K.,
Efendic, S., and Landau, B. R. (2001) Metabolism 50, 47–52
Sherry, A. D., Sumegi, B., Miller, B., Cottam, G. L., Gavva, S., Jones, J. G., and
Malloy, C. R. (1994) Biochemistry 33, 6268 – 6275
Cooper, A., Nieves, E., Rosenspire, K., Filc-DeRicco, S., Gelbard, A., and
Brusilow, S. (1988) J. Biol. Chem. 263, 12268 –12273
Williamson, J. R., Browning, E. T., and Scholz, R. (1969) J. Biol. Chem. 244,
4607– 4616
Hiltunen, J. K., and Hassinen, I. E. (1977) Int. J. Biochem. 8, 505–509
Reshef, L., Olswang, Y., Cassuto, H., Blum, B., Croniger, C. M., Kalhan, S. C.,
Tilghman, S. M., and Hanson, R. W. (2003) J. Biol. Chem. 278,
30413–30416
Kalhan, S. C., Mahajan, S., Burkett, E., Reshef, L., and Hanson, R. W. (2001)
J. Biol. Chem. 276, 12928 –12931
Beale, E. G., Hammer, R. E., Antoine, B., and Forest, C. (2004) Trends
Endocrinol. Metab. 15, 129 –135
Beale, E. G., Hammer, R. E., Antoine, B., and Forest, C. (2002) FASEB J. 16,
1695–1696
Eaton, S., Record, C. O., and Bartlett, K. (1997) Eur. J. Clin. Investig. 27,
719 –722
Lehninger, A., Nelson, D., and Cox, M. (2000) in Principles of Biochemistry,
(Nelson, D. L., and Cox, M. M., eds) 3rd Ed., pp. 586 –588, Worth Publishing, New York
Owen, O. E., Kalhan, S. C., and Hanson, R. W. (2002) J. Biol. Chem. 277,
30409 –30412
Younes, M., and Strubelt, O. (1988) Pharmacol. Toxicol. 63, 382–385
Beauvieux, M.-C., Tissier, P., Gin, H., Canioni, P., and Gallis, J.-L. (2001) J.
Nutr. 131, 1986 –1992
Dufour, S., Thiaudiere, E., Gallis, J.-L., Canioni, P., and Diolez, P. (1996)
Magn. Reson. Mat. Phys. Biol. Med. 4, 329
Scholz, R., Schwabe, U., and Soboll, S. (1984) Eur. J. Biochem. 141, 223–230
Gastaldelli, A., Baldi, S., Pettiti, M., Toschi, E., Camastra, S., Natali, A.,
Landau, B. R., and Ferrannini, E. (2000) Diabetes 49, 1367–1373
Sunehag, A. L., Treuth, M. S., Toffolo, G., Butte, N. F., Cobelli, C., Bier, D. M.,
and Haymond, M. W. (2001) Pediatr. Res. 50, 115–123
Sunehag, A. L., Haymond, M. W., Schanler, R. J., Reeds, P. J., and Bier, D. M.
(1999) Diabetes 48, 791– 800
Peroni, O., Large, V., and Beylot, M. (1995) Am. J. Physiol. 269, E516 –E523
Weis, B. C., Margolis, D., Burgess, S. C., Merritt, M. E., Wise, H., Sherry, A. D.,
and Malloy, C. R. (2004) Magn. Reson. Med. 51, 649 – 654