| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
regulates amino acid metabolism
Institut de Biologie Animale, Université de Lausanne, CH-1015, Lausanne, Switzerland;
* Pfizer Global Research and Development, Ann Arbor Laboratories, Ann Arbor, Michigan 48105, USA; and
Laboratory of Metabolism, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892, USA
1Correspondence: Nutrition, Metabolism and Genomics Group, Wageningen University, P.O. Box 8129, 6700 EV Wageningen, The Netherlands. E-mail: sander.kersten{at}staff.nutepi.wau.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
is a ligand-activated
transcription factor that plays an important role in the regulation of
lipid homeostasis. PPAR mediates the effects of fibrates, which are
potent hypolipidemic drugs, on gene expression. To better understand
the biological effects of fibrates and PPAR, we searched for genes
regulated by PPAR using oligonucleotide microarray and subtractive
hybridization. By comparing liver RNA from wild-type and PPAR null
mice, it was found that PPAR decreases the mRNA expression of
enzymes involved in the metabolism of amino acids. Further analysis by
Northern blot revealed that PPAR influences the expression of
several genes involved in trans- and deamination of amino acids, and
urea synthesis. Direct activation of PPAR using the synthetic
PPAR ligand WY14643 decreased mRNA levels of these genes, suggesting
that PPAR is directly implicated in the regulation of their
expression. Consistent with these data, plasma urea concentrations are
modulated by PPAR in vivo. It is concluded that in addition to
oxidation of fatty acids, PPAR also regulates metabolism of amino
acids in liver, indicating that PPAR is a key controller of
intermediary metabolism during fasting.—Kersten, S., Mandard, S.,
Escher, P., Gonzalez, F. J., Tafuri, S., Desvergne, B., Wahli, W.
The peroxisome proliferator-activated receptor regulates amino acid
metabolism.
Key Words: fasting • microarray • PPAR • SABRE
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
The molecular mode of action of PPARs is analogous to that of many
nuclear hormone receptors. They can be activated by certain ligands and
modulate DNA transcription by binding to specific nucleotide sequences
in the promoter of target genes. Three PPAR isotypes are known: ,
ß, and 
(NR1C1, NR1C2, NR1C3) (2)
. The PPAR
isotype is mostly expressed in liver and is the cellular target for
fibrates. Fibrates, which include gemfibrozil, bezafibrate, and
fenofibrate, are potent hypolipidemic drugs widely used to treat
cardiovascular disease. The PPAR isotype is mainly expressed in the
adipose tissue, where it promotes the differentiation of preadipocytes
into adipocytes. It is the target for a group of drugs called
thiazolidinediones, such as troglitazone and rosiglitazone, used to
treat type II diabetes. Last, PPARß is expressed in many tissues and
has been proposed to mediate the inhibitory effects of NSAIDs on
colorectal cancer (reviewed in ref 1
).
Of the three PPAR isotypes known to date, the PPAR isotype has been
comparatively well characterized. The list of known PPAR target
genes encompasses an array of genes that participate in various aspects
of lipid metabolism such as fatty acid uptake through membranes, fatty
acid binding in cells, fatty acid oxidation (both peroxisomal and
mitochondrial), and lipoprotein assembly and transport. Both positive
and negative target genes of PPAR are known. Many functional effects
of fibrates can be explained by the activation or repression of some of
these genes via PPAR (3)
.
Inasmuch as most of our knowledge about the importance of PPAR is
directly coupled to the function of its target genes, the
identification of genes regulated by PPAR may aid in further
deciphering the role of PPAR in vivo. In addition, the
identification of new PPAR targets may help us comprehend the
pharmacological effects of fibrates, whose action in vivo is only
partially understood.
In recent years, technologies aimed at identifying differentially
expressed genes have improved considerably, allowing for a systematic
comparison of gene expression between two or more samples. One of these
methods, entitled ‘selective amplification via biotin- and restriction
mediated enrichment’ (SABRE), has proved useful in identifying genes
that display a pronounced circadian rhythm in expression levels
(4)
. Another technique, called oligonucleotide microarray
(Affymetrix), permits the simultaneous expression monitoring of several
thousands of genes (5)
. This latter technique is
increasingly used to study overall changes in gene expression
(expression profiling) (6)
. To search for new genes
regulated by PPAR, we applied the SABRE and microarray technology to
the PPAR null mutant model by comparing liver mRNA of wild-type and
PPAR null mice (7)
. We show here that PPAR
influences the expression of numerous genes implicated in amino acid
metabolism.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
null mice on a sv129 background were used. Animal experiments were
approved by the Animal Authorization Commission of the Canton of Vaud
(Lausanne, Switzerland). Fasted animals were deprived of food for
24 h, starting at the beginning of the light cycle. WY14643 was
administered by dissolving it in the drinking water (0.1%) for 3 wk.
Oligonucleotide microarray
Total RNA was prepared from mouse livers using Trizol reagent
(Gibco BRL, Grand Island, NY); 10 µg of total liver RNA pooled from
four mice was used per oligonucleotide microarray experiment. cRNA
synthesis, hybridization, washing, and scanning were performed
according to standard Affymetrix protocols. Fluorometric data were
generated by Affymetrix Software and the gene chips were globally
scaled to all the probe sets with an identical target intensity value.
Affymetrix software translates the results of the hybridization into
numerical values (average difference, or Avg Diff). The Avg Diff
is a measure of the intensity of hybridization to perfect match
oligonucleotide probes compared with mismatch probes and is an
indicator of the absolute expression level of a gene. The data set was
filtered according to the following criteria. Only genes with an
absolute expression level (Avg Diff) above the threshold of 30 were
kept. All genes that showed a less than 1.5-fold difference in Avg Diff
values between wild-type and PPAR null mice were eliminated from the
analysis. Genes with a less than twofold difference in expression were
eliminated if the Avg Diff values were below 200. Genes with a
difference in expression between two- and threefold were eliminated if
the Avg Diff values were below 100.
SABRE
SABRE uses selective streptavidin-biotin affinity and
restriction enzyme site reconstitution to enrich for cDNA species more
abundant in one population than in another. One male sv129 wild-type
mouse and one male PPAR null mouse of 
8 months of age were killed
in the middle of the light cycle. The liver was sampled and total RNA
prepared using guanidinium thiocyanate. PolyA RNA was prepared using a
kit from New England Biolabs (Beverly, MA). mRNA was reverse
transcribed using a Promega cDNA synthesis kit (Madison, WI). cDNA was
digested with Sau3AI and ligated with adapters. Ligation products were
purified on low melting point agarose, taking a range from 150 bp to 1
kb. Products were amplified by PCR with primers complementary to the
adapter; 10 µg of driver PCR products was hybridized by PERT
hybridization with 0.33 µg of tester PCR products (0.33 µg driver
PCR products for control reaction). The hybridized solution was
digested with S1 nuclease and bound to streptavidin beads. Hybrids were
eluted with BamHI and used for a new round of PCR. For
details, see ref 4
.
The following primers were used:
primer T1: biotin-CCAGGATCCAACCGATC;
primer T2: biotin-CTGGGATCCAACCGATC;
primer D1: GGTCCATCCAACCGATC;
primer D2: CTGCCATCCAACCGATC.
Bands that were enriched in the wild-type vs. PPAR -/- selection
and PPAR -/- vs. wild-type selections were excised from the dried
denaturing gel and eluted by boiling in 40 µl
H2O for 10 min. The eluted fragments were
amplified by PCR and subcloned into a cloning vector for sequencing.
RNA preparation and Northern blots
Total RNA was prepared from frozen livers by RNeasy midikit
(Qiagen, Chatsworth, CA) or Trizol reagent (Gibco-BRL); 20–30 µg of
total RNA was loaded per lane. Electrophoresis, blotting, and
hybridization were according to standard protocols. DNA probes were
obtained by RT-PCR of mouse or rat liver total RNA and verified by
sequencing. The cDNA of the ribosomal protein L27 was used as a control
probe. Labeling was carried out by High Prime kit (Roche, Nutley, NJ).
Northern blots were quantified using an Elscript400 quantitative
scanner.
Metabolite assays
Blood was drawn by retro-orbital puncture. Blood was collected
into heparinized tubes, kept on ice, and spun for 10 min to collect
plasma. Plasma was kept at -70°C. Plasma urea and ammonia were
measured with a Sigma kit. Plasma beta-hydroxybutyrate was determined
using a Boehringer kit (Boehringer-Mannheim, Mannheim, Germany).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 6)
. To identify genes not yet known to be
regulated by PPAR, the mRNA population from livers of fasted
wild-type and PPAR null mice were compared using murine 6500 gene
oligonucleotide microarrays (Affymetrix). Fasted mice were used because
the effects of PPAR deletion are much more apparent during fasting
(8
9
10
11)
.
Of a total of 6519 genes, 2300 were found to be present according to
Affymetrix software. Using this software, the abundance of each gene is
scored as a numerical value (Avg Diff) that reflects the intensity of
hybridization to perfect match oligonucleotide probes compared with
mismatch probes. It is an indicator of the absolute expression level
and can be compared between two separate microarray hybridizations.
Genes whose Avg Diff values differed less than 1.5-fold between
wild-type and PPAR null mice were eliminated, as well as genes with
Avg Diff values below a certain threshold. This threshold was adjusted
based on the magnitude of the fold difference in expression between
wild-type and PPAR null mice (see Materials and Methods for
details).
To facilitate the interpretation of the data, analysis was confined to
a small subset of genes up-regulated in the livers of PPAR null
mice. Analysis of the remainder of the genes will be reserved for a
future publication.
Disruption of the PPAR gene resulted in up-regulation of genes
connected with amino acid metabolism and synthesis of amino
acid-derived products. This included, among others, arginase, cytosolic
aspartate aminotransferase, branched chain keto acid dehydrogenase, and
spermidine synthase (Table 1
). These data suggest that PPAR may decrease the expression of genes
involved in amino acid metabolism. Although a preliminary link between
fibrates and aspartate aminotransferase and alanine aminotransferase
has already been made (12)
, the overall effect of PPAR
on amino acid metabolism remains to be established. To confirm the
microarray data, Northern blots were performed for two genes: arginase
and cytosolic aspartate aminotransferase. Both genes were indeed more
highly expressed in fasted PPAR null mice than in fasted wild-type
mice (Fig. 1
). Quantitation of the Northern blot suggested that the oligonucleotide
microarray underestimated the fold difference in expression (3.8- vs.
11.6-fold for cytosolic aspartate aminotransferase and 1.6- vs.
2.1-fold for arginase).
|
|
Subtractive hybridization (SABRE)
In addition to microarray, a recently developed PCR-based
subtractive hybridization assay—SABRE (4)
—was used to
identify genes influenced by PPAR. Liver mRNA from one wild-type
sv129 and one PPAR null mouse was prepared. After several rounds of
subtractive hybridization, 12 cDNA fragments were isolated (Fig. 2
). Two of these cDNA clones corresponded to known PPAR target genes
(Cyp4A10=fatty acid 
hydroxylase and short chain acyl-CoA
dehydrogenase). One cDNA clone corresponded to arginino succinate lyase
and another to hydroxypyruvate/glyoxylate reductase, both of which are
implicated in amino acid metabolism. Northern blot confirmed their
differential expression between wild-type and PPAR null mice
(Fig. 3
). Whereas arginino succinate lyase was more highly expressed in PPAR
null mice, hydroxypyruvate/glyoxylate reductase mRNA was reduced in
these mice. These data provide additional evidence that disruption of
the PPAR gene influences the hepatic expression of genes involved in
amino acid metabolism.
|
|
It can be hypothesized that the elevated expression levels observed in
fasted PPAR null mice may be a consequence of the severe metabolic
perturbations present in these mice (9
, 10)
. To exclude
this possibility, the effect of direct ligand activation of PPAR was
studied by feeding mice the synthetic PPAR ligand WY14643
(Fig. 4
). Changes in expression greater than or equal to 1.5-fold or less than
or equal to 0.6-fold were considered significant. Administration of
WY14643 altered expression levels of cytosolic aspartate
aminotransferase, arginino succinate lyase, and
hydroxypyruvate/glyoxylate reductase in a PPAR dependent manner
and in an opposite direction from the changes observed in PPAR null
mice. Expression of arginase, which was increased in fasted and fed
PPAR null mice vs. control mice, was also increased by WY14643 (to
evaluate the effect of PPAR disruption in fed mice, compare lanes 1,
2 with lanes 5, 6). This suggests that regulation of arginase
expression by PPAR is more complex and contains both stimulatory and
inhibitory components. Taken together, it can be concluded that since
ligand activation of PPAR in normal mice results in altered
expression of cytosolic aspartate aminotransferase, arginino succinate
lyase, hydroxypyruvate/glyoxylate reductase, and arginase, the
differential expression between fasted wild-type and PPAR null mice
is probably not a consequence of metabolic perturbations accompanying
PPAR disruption.
|
PPAR influences expression of genes involved in transamination
and deamination
Amino acids are required for numerous processes, including protein
synthesis, energy generation, and the synthesis of specialized products
such as neurotransmitters and polyamines. Excess amino acids are either
transaminated or deaminated. Via a complex network of metabolic
reactions, the amino acid nitrogen is converted to urea, which is
excreted by the kidneys. The results presented above provide a
preliminary link between PPAR and amino acid metabolism. To more
closely examine the effect of PPAR on explicit pathways of amino
acid metabolism, mRNA levels of numerous enzymes were assessed by
Northern blot (Fig. 5
). The effect of ligand activation of PPAR was studied by
administering WY14643 and the effect of disruption of the PPAR gene
was studied by comparing fasted wild-type and PPAR null mice.
|
We first investigated the transamination process by determining the
expression of mitochondrial aspartate aminotransferase (expression of
cytosolic aspartate aminotransferase was shown above to be decreased by
PPAR). Expression of mitochondrial aspartate aminotransferase was
unaltered in fasted PPAR null mice and increased by WY14643 in a
PPAR-dependent manner. Alanine glyoxylate aminotransferase
expression was unaltered between fasted wild-type and PPAR null
mice, but was markedly decreased by WY14643 in a PPAR dependent
manner.
Oxidative deamination is mainly catalyzed by two enzymes, glutamate
dehydrogenase and glutaminase, which liberate ammonia. In the reverse
reaction, ammonia can be sequestered into glutamine by glutamine
synthase. Glutamine synthase expression was unaltered by WY1643 and in
fasted PPAR null mice, but mRNA levels were higher in fed PPAR
null mice than in fed wild-type mice (see right panel, compare lanes 1,
2 with lanes 5, 6). Expression of glutamate dehydrogenase was barely
higher in fasted PPAR null mice and was slightly decreased by
WY1643, but in a PPAR-independent manner. Expression of glutaminase
was much higher in fasted PPAR null mice than in wild-type mice and
was decreased by WY14643 in a PPAR-dependent manner. Together, the
data show that PPAR influences the expression of genes involved in
trans- and deamination of amino acids.
PPAR influences urea cycle gene expression
Deamination of amino acids in the liver releases ammonia, which is
efficiently converted to nontoxic urea by the urea cycle. The urea
cycle is catalyzed by five enzymes: arginino succinate lyase, arginino
succinate synthase, carbamoyl phosphate synthase 1, ornithine
transcarbamoylase, and arginase. The results of arginase and arginino
succinate lyase were shown above. The pattern of expression of arginino
succinate synthase greatly resembled that of arginino succinate lyase,
with higher expression of the gene in fasted PPAR null mice and
PPAR-dependent negative regulation by WY14643. A similar pattern was
observed for carbamoyl phosphate synthase 1, except that its expression
was elevated in fed PPAR null mice compared with fed wild-type mice.
Expression of ornithine transcarbamoylase was also down-regulated by
WY14643 in wild-type mice, but expression levels did not differ between
wild-type and PPAR null mice in the fasted state. Taken together,
the data indicate that PPAR has an inhibitory effect on the
expression of urea cycle enzymes in mice, with the exception of
arginase, whose regulation is more complex.
Plasma urea is increased in fasted PPAR null mice
The gene expression data indicate that PPAR has an overall
suppressive effect on the expression of several genes that participate
in ammonia and urea synthesis in the liver. To examine whether this has
any functional consequences on the rate of urea formation and release
in vivo, we measured the concentration of urea in plasma (Fig. 6A
). In the fed state, the plasma urea concentration was
identical between fed wild-type and PPAR null mice. In contrast,
plasma urea in the fasted state was significantly higher in PPAR
null mice than in wild-type mice. The increased plasma urea
concentration agrees with the increased mRNA levels of four out of five
urea cycle enzymes in fasted PPAR null mice vs. wild-type mice.
Apparently, changes in gene expression caused by PPAR are translated
into actual functional changes at the physiological level. No
differences were found in plasma ammonia concentration between
fasted wild-type and PPAR null mice (Fig. 6B
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
influences the
expression of numerous genes implicated in major pathways of amino acid
metabolism. This includes transamination (aspartate aminotransferase,
alanine glyoxylate transaminase), deamination (glutaminase), urea cycle
(all five enzymes), oxidation of alpha keto acids (branched chain keto
acid dehydrogenase, hydroxypyruvate/glyoxylate reductase), amino acid
inter-conversions (asparagine synthetase, phenylalanine hydroxylase),
and synthesis of amino acid-derived products (spermidine N1
acetyltransferase, spermidine synthetase). With the exception of
hydroxypyruvate/glyoxylate reductase and arginase, PPAR suppresses
the expression of these genes, leading to an overall decrease in amino
acid degradation. The functional importance of regulation of amino acid
metabolism by PPAR is substantiated by the demonstration that plasma
urea concentrations are increased in fasted PPAR null mice vs.
wild-type mice.
Differences exist between mice and humans with respect to their
metabolic response to fasting. Metabolism in mice is much more rapid.
Furthermore, whereas urea production in humans is increased during the
first 24–36 h of fasting, our data suggest that the opposite is true
in mice. However, there is no good reason to believe that the overall
effects of PPAR and PPAR ligands on genes involved in amino acid
metabolism would differ between mice and humans.
Contrary to common understanding, oxidation of amino acids contributes
significantly to energy production of several organs, including liver
and the gut (13)
. Amino acid oxidation is dramatically
elevated during conditions such as sepsis and cachexia and after severe
trauma and burns. These catabolic diseases are characterized by massive
net body protein breakdown, leading to a negative nitrogen balance.
Despite the clinical importance of amino acid metabolism and unlike
lipid and glucose metabolism, little information is available about its
regulation at the genetic level. It has been demonstrated that
glucocorticoids and glucagon increase expression of urea cycle enzymes
(14
15
16)
. Recent work has established the important role
of the transcription factor C/EBP in the regulation of urea
synthesis. Using C/EBP null mice, it was demonstrated that C/EBP
stimulates expression of urea cycle enzymes (17)
. Our data
point toward another global transcriptional regulator of the urea
cycle, PPAR, which by inhibiting mRNA expression of transamination,
deamination, and urea cycle enzymes has an effect directly opposite to
that of C/EBP. Consistent with these data, plasma urea levels are
increased in fasted PPAR null mice vs. wild-type mice.
PPAR activates gene transcription by binding to specific nucleotide
sequences in the promoter region of target genes. Positive targets of
PPAR include an array of genes that participate in various aspects
of fatty acid metabolism (1)
. In addition to stimulating
gene expression, PPAR also inhibits the expression of numerous genes
such as apoCIII (18
19
20)
, apoAI (in rodents) (20
, 21)
, transferrin (22)
, fibrinogen (23
, 24)
, pyruvate kinase (25)
, S14 (26)
,
Cyp2C family members (24)
, and positive acute-phase
response genes (24
, 27)
. Our microarray analysis revealed
that the number of genes up-regulated in livers of PPAR null mice
exceeds the number of genes that are down-regulated. Accordingly,
negative regulation by PPAR may be responsible for a plethora of
effects attributed to PPAR or to PPAR ligands, including its
effect on cell proliferation and inflammation (S. Kersten and P.
Escher, unpublished observations).
With respect to the mechanism of down-regulation by PPAR, it is not
unthinkable that in addition to a positive PPAR response element, there
may also be a negative PPAR response element that mediates repression
of DNA transcription by PPAR. However, such a promoter element has
yet to be identified. Alternatively, it is possible that PPAR
stimulates expression of another transcription factor that serves as a
transcriptional repressor. For example, down-regulation of the apoA-I
gene by PPAR in the rat has been proposed to be mediated by the
nuclear receptor rev-erb, whose expression is stimulated by PPAR
and is a negative regulator of apoA-I transcription. Rather than being
increased, however, we find that Rev-erb expression in mice is
slightly decreased by WY14643, in a PPAR-dependent manner (data not
shown). A final mechanism by which PPAR may repress transcription is
by competing with other transcription factors for a common coactivator
protein that is needed for transcriptional activation.
As fatty acids are ligands for PPAR, the suppressive effect of
PPAR on urea cycle enzymes may provide a potential explanation for
the inhibitory effect of fatty acids on ureagenesis (29)
and ammonia detoxification (30)
. Fatty acids have also
been shown to suppress arginino succinate synthase and carbamoyl
phosphate synthase expression in cell culture (31)
. This
mechanism may account for the abnormal expression of urea cycle enzymes
observed in carnitine-deficient juvenile visceral steatosis, a disease
characterized by defective fatty acid uptake into mitochondria and
associated accumulation of fatty acids in the cytosol
(32)
.
Why would the same transcription factor that stimulates hepatic fatty
acid oxidation suppress amino acid degradation and ureagenesis? During
prolonged fasting, fatty acid oxidation becomes the major source of
energy for the liver, an effect mediated by PPAR (9
, 10)
. At the same time, the relative contribution of amino acid
metabolism to hepatic ATP production, which is dominant in the fed
state (13)
, declines. In mice, this is associated with a
decreased expression of several amino acid metabolizing and
urea-synthesizing enzymes during fasting (S. Kersten et al.,
unpublished results). The reciprocal relationship between fatty acid
oxidation and nitrogen metabolism is illustrated by comparing the
plasma ketone body concentration, which reflects the rate of fatty acid
oxidation, and the plasma urea concentration, which (in the absence of
changes in renal clearance) is indicative of the rate of amino acid
metabolism and subsequent urea synthesis (Fig. 7
). It is conceivable that the simultaneous increase in ketone body
concentration and decrease in urea concentration during fasting in mice
are actually due to the action of a single factor, PPAR, which
balances the needs of the two pathways by altering the expression of
genes involved.
|
In addition to amino and fatty acid metabolism, recent evidence also
implicates PPAR in the regulation of carbohydrate metabolism. Using
the oligonucleotide microarray, it was shown that PPAR
up-regulates the expression of several genes involved in
gluconeogenesis. Overall, this suggests that PPAR acts as a global
regulator of energy metabolism in the liver, which coordinates the
rates of utilization of the various energy substrates in relation to
food availability.
In conclusion, it is shown here that fibrates and PPAR down-regulate
hepatic expression of genes involved in amino acid metabolism.
Regulation of amino acid metabolism by PPAR is supported by
functional data showing increased plasma urea concentrations in fasted
PPAR null mice. The data presented add to a growing body of evidence
indicating that negative regulation by PPAR may be responsible for
an abundance of effects attributed to PPAR or to PPAR
ligands.
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 7, 2001.
Revision received May 15, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
