The gene encoding rat 3-phosphoglycerate dehydrogenase
Emile Van Schaftingen
Laboratoire de Chimie Physiologique, Christian de Duve Institute of Cellular Pathology and Universite´ Catholique de Louvain, Avenue Hippocrate
75, B-1200 Brussels Belgium
Laboratoire de Biologie du De´veloppement, I.B.M.M., Universite´ Libre de Bruxelles, B-6041 Gosselies, Belgium
Received: 17 March 2000 / Accepted: 23 June 2000
The first step of
-serine synthesis is catalyzed by 3-phosphoglyc-
erate dehydrogenase (PHGDH), a ubiquitous enzyme in prokary-
otes and eukaryotes. The sequence encoding this enzyme, now
known for many species including rat (Achouri et al. 1997) and
human (Cho et al. 2000), shows that it belongs to the family of
-hydroxyacid dehydrogenases (Vinals et al. 1993), which also
comprises enzymes acting on
-glycerate and glycolate. In mam-
mals, PHGDH is widely distributed in tissues (Willis and Sallach
1964). However, its activity in rat liver, but not in other tissues,
strongly depends on the nutritional status, being low in animals fed
a normal diet and increasing more than 10-fold after shifting to a
protein-poor, carbohydrate-rich diet (Fallon et al. 1966; Mauron et
al. 1973). Previous work has shown that this control is exerted at
a pre-translational level by cysteine and glucagon, which appear to
act on the stability of the mRNA and on the transcription of the
gene, respectively (Achouri et al. 1999). PHGDH appears also to
be quite active in proliferating cells (Snell 1984), most likely ow-
ing to transcriptional regulation (Cho et al. 2000); this high activity
is probably related to the precursor role of serine in the biosyn-
thesis of nucleotides. Although the sequence of the cDNAs encod-
ing the rat and human enzymes has been published (Achouri et al.
1999; Cho et al. 2000), the structure of the corresponding gene is
not known for any metazoan species.
The purpose of this work was to determine the structure of the
gene encoding rat PHGDH and check whether the same promoter
was used in the liver as in tissues where no dietary regulation of
the expression is observed.
Screening of 10
plaques of a rat genomic library (from Strata-
gene, in Lambda Dash) with cDNA probes corresponding to the 5Ј
(nucleotides 1–514) and 3Ј (nucleotides 772–1780) ends of the rat
PHGDH cDNA (Achouri et al. 1997) yielded about 50 positive
clones. When digested with HindIII and analyzed by Southern
blotting, a majority of them showed one or two small (<4 kb)
bands hybridizing with both 5Ј and 3Ј probes derived from the
cDNA (not shown). Subcloning and sequencing (Achouri et al.
1997) showed that these clones contained intronless sequences
with multiple mutations compared with the cDNA (not shown),
allowing the conclusion that these sequences corresponded to
Southern blotting showed that other clones contained several
fragments that hybridized either with the 5Ј probe or with the 3Ј
probe (not shown). Their inserts were subcloned in pBluescript and
sequenced, allowing the identification of all exons except exons
6–8. The ≈4-kb region that was not covered by the clones was
obtained by PCR with primers corresponding to the missing exons
or intronic primers, and the genomic DNA library as template.
Except for a ≈250-bp, highly repetitive region in the second half of
intron 5, the whole gene and 5 kbp upstream of the 1
completely sequenced on both strands (≈20 kbp) or on one strand
with large overlaps between successive starting points (≈14 kbp).
Figure 1 shows the structure of the gene. It contains 13 exons,
one of which (exon 1Ј) is expressed in testis but not in other tissues
(see below). As shown in Table 1, all splicing sites conform to the
known consensus (Padgett et al. 1986). The introns range in size
from 0.1 to 10.5 kb, the largest one being between exons 5 and 6.
The structure of the gene markedly differs from those of Cae-
norhabditis elegans (accession number Z83219) and Arabidopsis
thaliana (AB010407.1), where only three exons are found. The rat
Phgdh gene and its 5Ј flanking sequence contain several dinucleo-
tide or trinucleotide repeats (e.g., a 24-TG repeat 380 bp upstream
of the initiator ATG), which could be helpful as polymorphic
RACE experiments were performed to determine the position
of the start site(s) in different tissues (liver, kidney, brain, lung,
and testis). The primers used for these experiments were derived
from exon 5 and exon 2, to allow the possible detection of an
exon. Sequencing of 24 clones derived from differ-
ent tissues (8 from liver and 4 from each of the four other tissues)
indicated that 10 of the RACE products started at position −21,
compared with the published cDNA sequence, 7 of them at posi-
tion −15, and 6 of them at different positions between −42 and +8
(Fig. 1), without any apparent tissue preference. These results
indicated that the Phgdh gene, like other housekeeping genes, has
no precise transcription start, in agreement with the lack of TATA
box. Furthermore, the promoter used in the liver did not appear to
be different from the one used in other tissues where the expres-
sion of the gene is not under dietary control.
One of the four RACE clones from testis started at position
−247, but in this clone, nucleotides −206 to +38 had been excised,
in agreement with the presence of splice site consensus sequences.
This alternative transcript has the same initiator ATG codon as the
known cDNA and encodes, therefore, the same protein. To check
the use of the more upstream start site, we analyzed total RNA
from different tissues by RT-PCR by using a sense primer in exon
1Ј (Fig. 1B) and an antisense primer in exon 2. PCR amplification
resulted in the formation of a product with the expected size (≈ 240
bp) only when it was performed with cDNA from testis. In con-
trast, amplification of a ≈270-bp fragment was observed with
cDNA from all tissues when a primer corresponding to exon 1
(shown in Fig. 1) was used together with the antisense primer in
exon 2. These results indicated that the more upstream start site
was used in testis but not in liver, kidney, brain, or lung.
The rat Phgdh gene was localized by using mouse × rat cell
hybrids, which segregate rat chromosomes (Szpirer et al. 1984).
The probe used, a PstI intronic fragment corresponding to nucleo-
tides 176–837 of the first intron and labelled by random priming,
Database accession number: RNO271975
Correspondence to: E. Van Schaftingen; E-mail: vanschaftingen@bchm.
© Springer-Verlag New York Inc. 2000Mammalian Genome 11, 1034–1036 (2000).