(Received for publication, October 25, 1996)
From the Lineberger Comprehensive Cancer Center and
Departments of § Pharmacology, and
Internal Medicine,
University of North Carolina School of Medicine,
Chapel Hill, North Carolina 27599
Catalysis of guanine nucleotide formation from
IMP in the de novo purine synthetic pathway is carried out
by two isoforms of the enzyme inosine monophosphate dehydrogenase
(IMPDH) that are catalytically indistinguishable but are encoded by
separate genes. In order to assess the potential for cell type-specific expression of IMPDH activity, we have characterized the IMPDH type I
gene and identified three major RNA transcripts that are differentially
expressed from three different promoters. A 4.0-kilobase pair (kb)
mRNA containing 1.3 kb of 5-untranslated region is expressed in
activated peripheral blood lymphocytes and to a far lesser extent in
cultured tumor cell lines. The P1 promoter that regulates the
transcription of this mRNA has a high degree of sequence identity
to an Alu repetitive sequence. A transcript of 2.7 kb is found in a
subset of the tumor cell lines examined, whereas a 2.5-kb mRNA
species is universally expressed and is the prevalent mRNA in most
cell lines and tissues. The relative strengths of the three promoter
regions and the effects of variable extents of 5
-flanking sequence on
the P3 promoter differ in Jurkat T, as compared with Raji B lymphoid
cell lines, demonstrating a complex cell type-specific transcriptional
regulation of IMPDH type I gene expression.
Inosine 5-monophosphate dehydrogenase
(IMPDH;1 EC 1.1.1.205) is an essential,
rate-limiting enzyme in the de novo guanine nucleotide
synthetic pathway. It catalyzes the conversion of IMP to XMP at the
purine metabolic branch point and provides sufficient guanine
nucleotides for important cellular processes. The observations that the
activity of IMPDH is tightly linked with both cellular proliferation
and transformation (1-3) have led to an interest in developing IMPDH
inhibitors for clinical use. Inhibition of IMPDH activity is associated
with depletion of intracellular guanine nucleotide pools and has been
shown both to inhibit cellular proliferation (4, 5) and to induce cell
differentiation (6-9). The specific reversibility by exogenous guanine
of the biologic effects of IMPDH inhibitors on cell division and
differentiation is a unique feature of such compounds that
differentiates them from other inducers of cellular differentiation.
IMPDH inhibitors have also been demonstrated to have considerable
efficacy as immunosuppressive agents (10, 11) and prevent both B and T
lymphocyte activation (12, 13).
Human IMPDH activity is composed of the activities of two separate but very closely related IMPDH isoenzymes, termed type I and type II, that are encoded by separate genes located on chromosomes 7 and 3, respectively (14, 15). The corresponding cDNAs have been isolated (16, 17) and encode two distinct protein subunits of 56 kDa with 84% sequence identity. The two IMPDH proteins are tetrameric and are indistinguishable in their catalytic activities, substrate affinities, and Ki values for known inhibitors (18-20).
The increased IMPDH activity observed in replicating or neoplastic cells is largely due to increased expression of the type II IMPDH mRNA, whereas expression of the type I mRNA has been thought to be relatively unchanged by cell proliferation or transformation (21, 22). Conversely, type II mRNA levels are sharply decreased in response to the induction of cell differentiation, whereas type I mRNA remains constitutively expressed (22). A survey of relative IMPDH mRNA levels in human tissues demonstrated significant differences in the pattern of distribution of the type I transcript, with relatively high levels in kidney, pancreas, colon, and peripheral blood leukocytes as well as in fetal heart, brain, and kidney (23). Expression levels of the type II transcript, while generally higher than type I, were far less variable in different tissue types. These data support the concept that the two genes are regulated differently in resting, proliferating, and neoplastic cells and that these differences may have important consequences for cellular function.
In view of the important role of IMPDH activity in the immune response, we have previously studied IMPDH type I and II expression during T lymphocyte activation and documented that IMPDH type I mRNA levels increase approximately 10-fold in response to a variety of stimuli. In addition, T cell activation results in the appearance of two distinct type I mRNA transcripts, the larger of which is not readily appreciated in Northern blots of RNA from other cell types (24). This heterogeneity in IMPDH type I expression has led us to characterize the IMPDH type I gene and compare it with the recently analyzed IMPDH type II gene (25) with the goal of elucidating possible mechanisms involved in the tissue-specific expression of the IMPDH type I gene.
The cloning
of the IMPDH type I gene was complicated by the presence in the human
genome of multiple processed pseudogenes (26). To obtain IMPDH I
genomic clones, a human leukocyte genomic library in the phage FIX
II vector (provided by Dr. J. Lowe, University of Michigan, Ann Arbor)
was sequentially screened using oligonucleotides complimentary to the
5
- and 3
-untranslated regions (UTRs) of the type I cDNA (17). Two
positive clones were isolated from the approximately 1.8 × 106 plaques screened. Clone 1111a was obtained by
sequential hybridization to three oligonucleotides corresponding to the
5
-UTR of the cDNA (bp
85 to
108,
523 to
546 and
448 to
471, relative to the ATG). Clone 1711a was obtained using two
oligonucleotides corresponding to the 3
-UTR (bp 1614-1637 and
2119-2142). The identity of each clone was confirmed by hybridization
with the cDNA of the entire coding region. The probes were either
end-labeled with [
-32P]ATP (6000 Ci/mmol; Amersham
Corp.) using T4 polynucleotide kinase (Promega, Madison, WI) or labeled
using a random primer labeling kit (Promega) with
[
-32P]dCTP (3000 Ci/mmol; Amersham) to a specific
activity of 5-10 × 108 cpm/µg. Hybridizations were
performed in a 10% dextran sulfate, 1% SDS, and 5.8% NaCl solution
at 60 °C for 16-24 h, and the blots were washed twice with 1 × SSC, 0.2% SDS at room temperature for 10 min, and once at
50-55 °C for 30 min for oligonucleotide probes. Hybridization using
the cDNA probe was performed at 65 °C and with a final wash in
0.1 × SSC, 0.1% SDS at 65 °C for 30 min. Both clones were
digested with the restriction endonuclease SacI and the
resulting fragments subcloned into pGEM7Zf(+) vector. Mapping of the
gene was performed using Southern blot and PCR analysis. The entire
coding region and all exon-intron boundaries were sequenced using the
dideoxynucleotide chain termination method (27) and Sequenase 2.0 (U.S.
Biochemical Corp.). For the promoter regions, both automated
double-stranded sequencing and sequencing of single-stranded DNA were
performed in regions with high GC content.
Total cellular RNA was isolated using TRI-reagent (MRC, Cincinnati, OH). Total cellular RNA (15 or 20 µg) was electrophoresed on 1% agarose, 2.2 M formaldehyde gels and transferred to Zeta-Probe GT membranes (Bio-Rad). Prehybridizations were performed in 1% SDS, 0.1 M NaCl for 1 h at room temperature. Hybridizations were carried out in high efficiency hybridization solution containing 50% formamide (MRC) with 32P-labeled probe at 2-3 × 106 cpm/ml at 42 °C for 24-48 h. Oligonucleotide probes were washed in 1 × SSC, 0.2% SDS at 50 °C for 30 min, the p190 probe was washed in 0.1 × SSC, 0.1% SDS at 50-60 °C for 30 min, and the p700 and 1.5-kb cDNA probes were washed in 0.1 × SSC, 0.1% SDS at 65 °C for 30 min.
Human peripheral blood lymphocytes (PBLs) were isolated from buffy coats obtained from the American Red Cross (Charlotte, NC) using Histopaque-1077 (Sigma) gradient centrifugation. Monocytes and macrophages were removed using the adherence method (28). Purified PBLs were incubated at a concentration of 1.5 × 106/ml in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum (Life Technologies, Inc., Grand Island, NY), 100 units/ml penicillin, and 100 µg/ml streptomycin for 24 h in the presence or absence of 10 ng/ml of phorbol 12-myristate 13-acetate and 250 ng/ml of ionomycin (Calbiochem). The extent of lymphocyte activation was monitored by measuring [3H]thymidine incorporation into DNA at 24 and 48 h of activation.
Primer ExtensionAn oligonucleotide primer complimentary to
bases 73-99 in exon 1 was end-labeled with [-32P]ATP
using T4 polynucleotide kinase. Probe (1 × 105 cpm)
was mixed with 20 µg of total RNA from Jurkat and Raji cells or tRNA
as a control and precipitated. Annealing was performed in 20 µl of
1 × avian myeloblastosis virus reverse transcription buffer
(Promega) and incubated sequentially at 75 °C for 15 min and
60 °C for 30 min and slowly cooled to 40 °C. Extension was performed in the presence of 20 units of avian myeloblastosis virus
reverse transcriptase (Promega), 1 mM dNTPs, and 10 units of RNasin (Promega) at 42 °C for 90 min. The mixture was then digested with 20 µg/ml DNase-free RNase (Boehringer Mannheim) at
37 °C for 20 min followed by phenol/chloroform extraction and ethanol precipitation. The resulting products were resuspended in 5 µl of gel loading buffer, and 3 µl was analyzed on a denaturing 8 M urea, 6% polyacrylamide gel. The gel was dried and
autoradiographed at
70 °C for 2 days.
A pGEM-320 construct,
containing genomic DNA extending 280 bp upstream and 45 bp downstream
of the ATG translation initiation site was used for RNase protection
assays using the Ribonuclease Protection Assay kit (Ambion, Austin,
TX). The construct was digested with the restriction endonuclease
BglI, generating a 1.6-kb fragment containing the SP6
promoter and 115 bp 5 to the ATG initiation codon. This fragment was
gel-purified and transcribed in vitro using SP6 RNA
polymerase (Promega) to generate a [32P]CTP-labeled
235-bp antisense RNA probe. The labeled probe (5 × 105 cpm) was added to 30 µg of total RNA from Jurkat and
Raji cells and 30 µg of tRNA, and ethanol-precipitated.
Hybridizations were performed in 20 µl of hybridization buffer at
45 °C for 20 h, and samples were digested with a RNase A and T
mixture (1:100 dilution) at 37 °C for 30 min and precipitated in the
presence of 2 µg of tRNA. The pellets were resuspended in 5 µl of
gel loading buffer, and 3 µl was analyzed on 6% denaturing
polyacrylamide gels.
The P1 and the P2 promoter regions were
obtained by PCR using the following primers: 32F
(5-TGTAATCCCAGCATTTTGGGA-3
) and 33R (5
-GCTCTGTCGCCCAGGCTGGAGT-3
)
for the P1 region; 7F (5
-TTCTTTCCAGTCCCACCCGTGTAG-3
) and 18R
(5
-TGGCGTTTCGGGAAGTTA-3
) for the P2 region (see Fig. 3). The
amplified regions were cloned into the PCRTM II vector
(Invitrogen, San Diego, CA) and subcloned into the HindIII
site of the pCAT-Basic vector (pCATB) in 5
3
orientation upstream
of the chloramphenicol acetyltransferase (CAT) gene. Fragments of 0.4, 0.7, 1.4, and 2.7 kb 5
to exon 1 were isolated by restriction enzyme
digestions, as indicated in Fig. 6, gel-purified, and subcloned into
the SalI site of pCATB. The pCATB/0.7 construct is
designated as P3. The orientation of each construct was verified by
direct sequencing.
Sequences of the IMPDH type I P1 (A), P2 (B), and P3 (C) promoters. The three putative promoter regions used to assay CAT activity are shown in boldface type. Putative transcription factor binding sites based on consensus sequences are indicated by underlining. The primers (32F, 33R, 7F, and 18R) used for PCR are indicated. Transcription initiation sites based on primer extension analysis and confirmed by RNase protection assays are shown by arrows. The transcription start sites determined by primer extension only are shown by arrows with parenthesis. The P3 sequence is numbered relative to the A of the ATG start codon (+1) in exon 1.
Transient Transfection and CAT Assays
Plasmid DNA used for
transfection experiments was prepared using QIAGEN columns
(Chatsworth, CA), and the DNA was quantitated by both
spectrophotometry and ethidium bromide staining. To determine the
transcriptional activity of the putative promoters, 20 µg of each
construct was added to 107 exponentially growing Jurkat or
Raji lymphoblasts in a total volume of 500 µl of serum-free RPMI 1640 medium. Electroporations were performed using a Bio-Rad Gene Pulser set
at 250 V and 960 microfarads. Cells were plated in 20 ml of RPMI 1640 medium containing 10% fetal bovine serum, 100 units/ml of penicillin,
and 100 µg/ml of streptomycin and cultured for 44-48 h at 37 °C
in a humidified atmosphere in the presence of 5% CO2. A
-galactosidase plasmid under control of the
-actin promoter
(p
Ac-lacZ) (kindly provided by Lorraine Gudas, Cornell University)
was used in each experiment to control for transfection efficiency.
After harvesting, cells were washed twice in cold phosphate-buffered
saline and extracted in 150 µl of 0.25 M Tris-HCl (pH
8.0) with three cycles of freeze-thawing. After centrifugation at
14,000 × g for 5 min, supernatants were used for
protein quantitation using the Micro BCA protein assay reagent kit
(Pierce) and for assay of
-galactosidase activity by a modification
of a published technique (29) using chlorophenolred
-D-galactopyranoside (Boehringer Mannheim) as a
substrate. For CAT assays, aliquots of supernatant were heated at
60 °C for 10 min followed by centrifugation at 14,000 × g for 5 min, and 50 µl of supernatant were then incubated
with 0.125 µCi of [14C]chloramphenicol (Amersham) and
25 µg of n-butyryl coenzyme A (Sigma) for
2-6 h, depending on the transfection efficiency of each experiment.
Reactions were extracted with xylenes and counted in a Beckman
scintillation counter (model LS 7800; Beckman, Irvine, CA). CAT
activity was normalized for extract protein concentration and
calculated as cpm/mg/min of assay.
Two phage clones
(1111a (14 kb) and 1711a (13 kb)) were isolated that overlap by 3.6 kb
in the intragenic region and extend approximately 6.3 kb 5 and 4.7 kb
3
from the ends of the coding region. A physical map of the gene was
established using a combination of restriction mapping, PCR, and
sequence analysis of the two genomic clones (Fig. 1).
The entire IMPDH type I gene is approximately 18 kb in length.
According to the published cDNA sequence (17), the 5
-UTR is
composed of 600 bp, which we have designated as exons A-C in Fig. 1.
Exon A
is the region between exon A and an upstream transcription
initiation site described below. The coding region of the gene contains
14 exons, the last of which includes 713 bp of 3
-UTR. All exon-intron
boundaries contain the canonical splice donor (AG) and acceptor (GT)
consensus sequences, with the exception of a GC splice acceptor site at
exon 4. Within the coding region, 16 bases differ from the published
cDNA sequence (17). Six of these base changes result in amino acid
differences: bp 86 (A
G; aspartic acid
glycine); 327 (C
G;
asparagine
lysine); 819 (C
G; phenylalanine
leucine); 820 (C
G; histidine
aspartic acid); 1255 (C
G; proline
alanine); and 1489 (C
G; proline
alanine).
Comparison of the Genomic Structures of the IMPDH Type I and II Genes
The type I gene is significantly larger than the 5.8-kb
type II gene (25). However, as shown in Table I, the
exons of the coding region are identical in size and contain the same
amino acids at each exon boundary with the exception of an isoleucine valine substitution at the exon 5/6 boundary. The type I gene contains substantially larger introns, seven of which are over 1 kb in
length, and has approximately 1.3 kb of 5
-UTR and 713 bp of 3
-UTR, as
compared with 50 and 53 bp in the type II gene, respectively (16, 25).
In addition, nine ATG codons are present in the 5
-UTR of the type I
gene.
|
Previous studies have
demonstrated two IMPDH type I mRNA species of approximately 3 and 4 kb on Northern blot analysis of RNA from peripheral blood T lymphocytes
(24). RNase protection assays suggested that the larger mRNA
species of 4 kb contained additional 5 or 3
sequences, since the
coding regions of both mRNA species were identical. In order to
determine the composition of these mRNAs, Northern blot
hybridization was performed on total cellular RNA isolated from resting
or activated human PBLs using probes corresponding to different exons.
Using a cDNA probe corresponding to the coding region (exons
1-14), the expression of both a 4- and a 2.5-kb RNA species, as
accurately assessed using RNA markers, was markedly induced as a result
of phorbol 12-myristate 13-acetate plus ionomycin treatment of PBLs
(Fig. 2A, lanes 1 and
2). When a 700-bp probe encompassing 500 bp in exon A
and
200 bp in exon A (p700; Fig. 2A, lanes 3 and
4) and a 190-bp probe spanning the region from exon B to
exon C (p190; Fig. 2A, lanes 5 and 6)
were used as probes, only the 4-kb mRNA was observed, indicating
that this mRNA species originates upstream of the 2.5-kb mRNA
and contains a significantly longer 5
-UTR. In contrast, both mRNA
species were readily detected by two oligonucleotides derived from exon 1 sequence (Fig. 2A, lanes 7 and 8).
Northern blot analysis was also performed on total cellular RNA
isolated from nine tumor or transformed cell lines: an erythroleukemia
line, K562; a monocytic leukemia line, U937; three T cell leukemia cell
lines, KT-1, MOLT-4, and Jurkat; three B lymphoblast cell lines, MGL-8,
Nalm-6, and Raji; and a lung cancer cell line, H1437 (Fig.
2B, lanes 1-9). Upon hybridization with the p190
probe, as shown in Fig. 2B, a third RNA species of
approximately 2.7 kb was detected with relatively high abundance in
U937, Jurkat, and H1437 cells but was absent in MGL-8 and Raji cells,
as well as in PBLs. In order to confirm that this 2.7-kb band is not
identical to the major 2.5-kb mRNA transcript, the blot was
rehybridized with the 1.5-kb cDNA coding region probe. As shown in
Fig. 2C, all of the cell lines demonstrated an abundant
2.5-kb transcript, with U937 and Raji cells having the highest level of
expression after normalized for 28 S ribosomal RNA (Fig.
2D). These results clearly indicate the existence of three
IMPDH type I RNA transcripts that differ at their 5
termini.
Identification and Functional Analysis of the IMPDH Type I Promoter Regions
To study the elements involved in the regulation of IMPDH
type I transcription, approximately 2.7 kb 5 of exon C and 700 bp 5
of exon 1 were sequenced. Three regions were identified as putative
promoter regions based on sequence analysis; each contains a number of
putative transcription factor binding sites, and we have designated
them as P1, P2, and P3, respectively (Fig. 1). As shown in Fig.
3A, the P1 region encompasses 245 bp, as defined by PCR primers 32F and 33R, is located immediately 5
to exon
A
, and includes five Sp1 and two AP2 sites, one serum response element
site, one NF
B site that overlaps with a SIF consensus binding
sequences, and one ELP site. There is no TATA box in this region,
although a consensus sequence for transcription factor IID is present.
Of interest is the fact that this sequence is 95% identical to the
catarrhine-specific (CS) Alu repetitive sequence found widely
throughout the genome (30), although the consensus Alu sequence does
not function as a promoter in similar transfection experiments (data
not shown). A highly AT-rich region of 52 bp, containing four AAAAT
repeats plus stretches of A nucleotides separated by a single T or C is
located 3
to the P1 element. Primer extension analysis using RNA from
both resting and activated PBLs demonstrated two possible transcription
initiation sites downstream of the P1 promoter, as indicated in Fig.
3A. Ribonuclease protection analysis detected a single band
corresponding to the more 3
initiation site, although the
absence of a longer protected fragment could have resulted from
nonspecific cleavage by RNase A of the A-rich sequence (data not
shown).
The P2 region (Fig. 3B), 595 bp in length, is located 118 bp
5 to exon B, and contains 73% G + C residues, a number of Sp1 and AP2
binding sites, a PuF consensus binding site, and an Ets-1 consensus
binding site. A putative transcription initiation site located 134 bp
5
to exon B was determined by primer extension analysis using a primer
located in exon C. The P3 region (Fig. 3C), 671 bp in
length, is located immediately 5
to exon 1 and also has a high G + C
content of 73%. This region contains, in addition to a number of
potential Sp1 and AP2 binding sites, three Egr-1 sites (two of them
overlapping with Egr-2 sites), three PuF, two serum response elements,
and a site for CCAAT/enhancer-binding protein (C/EBP), a cAMP-response
element (CRE), and a CCAAT binding site. Primer extension was performed
using total RNA isolated from Jurkat and Raji cell lines and an
oligonucleotide corresponding to the exon 1 region. Two transcription
initiation sites were identified 40 bp (major) and 24 bp (minor) 5
to
the ATG initiation codon (Fig. 4A), and these
results were confirmed by a ribonuclease protection assay in which two
corresponding protected fragments were detected using RNA from Jurkat
and Raji cells but not with the tRNA control (Fig. 4B). The
greater intensity of the primer-extended bands and RNase-protected
bands in Raji cells is consistent with the higher levels of the 2.5-kb
mRNA in this cell line (Fig. 2C).
Functional studies of the P1, P2, and P3 regions were performed using
transient transfection assays with CAT reporter gene constructs (Fig.
5). In both Jurkat and Raji cell lines, the activity of
the P3 region is greater than that of the P1 and P2 regions. In the
Jurkat T cells, constructs containing P1 or P2 result in an approximate
20-fold increase in CAT activity over the values obtained with the
pCATB vector alone. In the Raji B cell line, in contrast, the P1 and P2
constructs resulted in a 1.2-6-fold increase in CAT activity, although
P3 retained a level of promoter activity equivalent to that found in
Jurkat cells. In order to define the P3 promoter region further,
additional constructs containing variable regions 5 to exon 1 were
compared for their ability to promote transcription in Jurkat and Raji
cells (Fig. 6). In Jurkat cells (Fig. 6A),
the construct pCATB/0.7, containing a 671-bp
ApaI-NotI fragment of P3, had the strongest
promoter activity, while the activity of construct pCATB/2.7,
containing additional 5
sequences, is only about 15-20% that of the
pCATB/0.7 construct. This profile differs markedly from that in Raji
cells (Fig. 6B), where pCATB/2.7 has approximately 6-fold
higher promoter activity than the pCATB/0.7 construct. These results
strongly suggest that there are elements 5
to the P3 region that may
regulate transcription from this promoter in a cell type-specific
fashion.
Transcriptional activity from the P3 element (pCATB/0.7) was compared with that from the 461-bp promoter for the IMPDH type II gene (25). CAT activity obtained with the type II promoter was 70-fold greater than that with the P3 promoter in both Jurkat and Raji cells (data not shown). This result is consistent with the observation that the expression of the type II IMPDH gene at the mRNA level is significantly higher than that of the type I gene in most tissues and cell lines tested (23) and suggests that transcriptional regulation may play a major role in determining the relative levels of expression of these two genes.
The co-existence in mammalian cells of two IMPDH genes with striking conservation of amino acid identity and catalytic function argues that each subserves an important and nonoverlapping function in cellular physiology or development. Since the expression of the IMPDH type II gene at the mRNA level has correlated more directly with cellular proliferation and transformation, it has been assumed that the IMPDH type I gene might provide a constitutive and noninducible level of guanine nucleotide biosynthesis (21, 22). However, the striking increase in IMPDH type I mRNA expression with T lymphocyte activation, the presence of a second mRNA transcript in these cells, and the variability in IMPDH type I gene expression in different tissues all support the concept that the regulation of expression of this gene may play a more significant role in cellular responses than had previously been appreciated. It was for this reason that a study of the structural characteristics of the IMPDH type I gene was undertaken.
The finding that the IMPDH type I gene retains the same exon structure
as the smaller IMPDH type II gene argues for an early gene duplication
event, while the preservation of the two functional genes throughout
vertebrate evolution (26) supports an essential role for each. There
are other striking examples of multigene families that account for
tissue-specific differences in the expression of human enzymes
important for nucleotide metabolism. Phosphoribosyl pyrophosphate
synthetase (EC 2.7.6.1) catalyzes the synthesis of phosphoribosyl
pyrophosphate, the primary precursor of the ribose monophosphate moiety
of all nucleotides and a critical regulator of de novo
purine and pyrimidine biosynthesis (31). Two isoforms,
phosphoribosylpyrophosphate synthetase subunits I and II, are encoded
by distinct genes (PRPS1 and PRPS2) located on
different regions of the X chromosome (32), and are expressed in a
tissue-specific manner (33). A third phosphoribosyl pyrophosphate synthetase gene encoding a testes-specific isoform is autosomal in
location (34). The tissue specificity of the expression of these genes
has been attributed to the differences in 5-flanking sequences and
corresponding differences in transcriptional activity (35). Similarly,
there are four isoforms of AMP deaminase (EC 3.5.4.6) encoded by three
AMP deaminase genes (36). Recent studies have demonstrated alternative
transcripts from the AMPD2 gene that are expressed in a
mutually exclusive pattern and that confer variable N-terminal
extensions to their encoded peptides (37), while multiple RNA
transcripts from the AMPD3 gene are regulated by three
distinct promoters in the 5
-flanking region (38). It appears from
these and other examples that the tissue-, development-, and
growth-regulated expressions of genes encoding enzymes essential for
nucleotide metabolism may have important consequences for the
development and survival of the organism as a whole.
The regulation of IMPDH type II gene expression has been studied in
detail in resting and activated T lymphocytes and has been shown to be
transcriptional in nature (25).2
Stimulation of T cells with phytohemagglutinin plus interleukin-2 induces increased transcription from a single core promoter region that
contains tandem cAMP-response element binding sites, an Sp1 site, and a
novel palindromic octamer sequence.2 In contrast, the
regulation of IMPDH type I expression is more complex and is governed
by the presence of three alternative promoters in the 5-flanking
region (Fig. 7). The P1 promoter gives rise to a 4.0-kb
transcript that is primarily found in PBLs, and expression of this
transcript increases markedly with T cell activation, as does
expression of the 2.5-kb mRNA from the more generally utilized P3
promoter. The presence of an NF
B consensus binding site within the
P1 region may be of direct relevance to this observation, since NF
B
is involved in mediating the transcription of a number of genes
following B and T lymphocyte activation, as well as in response to
inflammatory cytokines (40). The 10-fold increase in IMPDH enzymatic
activity and consequent increase in guanine nucleotide biosynthesis
that occurs during T lymphocyte activation (5, 24) are prerequisites
that enable these cells to enter S phase, and these events appear to be
extremely important in T cell biological responses (13, 41).
Although the relative contributions of IMPDH type I and II
proteins to this increase in activity have been difficult to determine,
given their high degree of amino acid identity, it is possible that the
appropriate regulation of IMPDH type I gene expression is essential for
the T cell mitogenic response.
The long 5-UTR found initially in the published cDNA sequence (17)
and demonstrated more fully as a result of our genomic characterization
may play an additional role in regulating IMPDH type I activity. In the
4-kb IMPDH type I RNA transcript, there are 1.3 kb of sequence 5
to
the known translation start site that include nine AUG codons followed
by open reading frames encoding putative peptides ranging from 3 to 85 amino acids. Two of these upstream AUG codons are in frame with the
initiation codon for methionine in exon 1, raising the possibility that
IMPDH type I isoforms with an additional 85 or 8 amino acids could be
generated. Although we have been unable to identify size variants of
IMPDH type I protein in Western blots of activated PBL extracts (data not shown), it is certainly possible that protein with N-terminal extensions could be synthesized. In addition, there is increasing evidence that 5
-UTR sequences longer than several hundred nucleotides may, as a consequence of their propensity to form secondary structure, regulate gene expression at the translational level (42). Although the
presence of multiple AUG translation start codons have been shown to
attenuate translation (43), there are also precedents for positive
regulation, as in the induction of yeast GCN4 translation in response
to amino acid availability (44) and in the Rous sarcoma virus RNA
translation (45). Furthermore, tissue specificity of expression may be
dictated by translational control mediated through this region (46,
47). Thus, although the 4-kb transcript is present in relatively low
abundance in the tissues other than PBLs that have been surveyed to
date, we cannot preclude a significant biologic role for this RNA
species in regulating IMPDH type I enzyme synthesis.
The P2 promoter is GC-rich and appears to regulate the expression of a 2.7-kb transcript in a manner that is disproportionate to the expression of the 2.5-kb transcript in several cultured cell lines. Since this band can only be clearly dissociated from the 2.5-kb transcript by differential hybridization to the p190 probe, it is difficult to determine the extent to which it is present in the tissues that have been studied to date (23). Although the structure of the P2 region does not appear to contain transcription factor binding sites that are highly tissue-specific in nature, the observation that the relative expression of CAT from a reporter construct containing 595 bp of sequence from this region is significantly higher in Jurkat that in Raji cells, together with the presence of the 2.7 transcript in the former but not the latter cell line, would support the concept of tissue-specific expression.
The P3 promoter accounts for the major 2.5-kb transcript and contains a large number of Sp1 and AP2 binding sites that may be responsible for the basal expression of this gene. The presence of several NGFI-A/Egr-1 sites in close proximity to the transcription initiation sites offers one potential explanation for the up-regulation of the 2.5-kb mRNA with T cell activation, since Egr-1 is a zinc finger protein induced following interleukin-2 stimulation of T lymphocytes that appears to be important for subsequent cellular proliferation (48). In addition, the presence of three consensus binding sites for PuF further upstream raises the possibility that NM23.H2, a product of one of two NM23 genes, could be involved in IMPDH I regulation. NM23.H2 is one of two subunits of nucleoside diphosphate kinase (EC 2.7.4.6), an enzyme that is important both in maintaining nucleoside triphosphate pools (49, 50) and in activating GTP-dependent proteins (51, 52). NM23.H2 has been shown to function as a transcription factor for the c-myc gene (53) and, together with NM23.H1, to be up-regulated in human PBLs following phytohemagglutinin stimulation with a time course that is consistent with the increase in IMPDH type I mRNA expression (24, 54). Whether or not these specific binding sites are important for the up-regulation of the 2.5-kb transcript in PBLs remains to be determined.
Of additional interest with regard to the P3 promoter region is the
differential expression of CAT constructs containing variable amounts
of 5 sequence extending as far as the SacI site 3
to exon
C (Fig. 1). Tissue specificity is evident in that maximal activity in
Jurkat T cells is obtained with the 700-bp region 5
to the ATG that
includes the PuF and Egr-1 sites, whereas 6-fold greater relative
activity is found in Raji cells when an additional 2 kb of upstream
sequence is included in the construct. This observation again
underscores the potential complexity of IMPDH type I gene transcriptional regulation. In contrast to the concept that the expression of IMPDH type I gene does not vary as a function of cell
proliferation or transformation, our results support the view that the
IMPDH type I gene is subject to differential regulation at the
transcriptional level from three promoter regions in a highly tissue-
or cell-specific manner. The list of enzymes whose expression is
regulated by multiple promoters and alternative processing at the
5
-end, as well as by the presence of multiple genes with varying
regulatory sequences encoding nearly identical proteins, is growing
steadily (39). The biological roles of IMPDH type I that require this
diversity of regulation remain to be determined.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) Y08944[GenBank], Y08945[GenBank], and Y08946[GenBank].