| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Center for Reproductive Sciences Department of Obstetrics & Gynecology & Reproductive Sciences and The Metabolic Research Unit University of California San Francisco, California 94143-0556
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-aminobutryic acid (GABAA) and
N-methyl-D-aspartate (NMDA) receptors
and also influence neuronal morphology and function.
Neurosteroids are produced in the brain using many of the same enzymes
found in the adrenal and gonad. The crucial enzyme for the synthesis of
DHEA (dehydroepiandrosterone) in the brain is cytochrome
P450c17. The transcriptional strategy for the expression of P450c17 is
clearly different in the brain from that in the adrenal or gonad. We
previously characterized a novel transcriptional regulator from Leydig
MA-10 cells, termed StF-IT-1, that binds at bases -447/-399 of the
rat P450c17 promoter, along with the known transcription factors
COUP-TF (chicken ovalbumin upstream promoter transcription factor),
NGF-IB (nerve growth factor inducible protein B), and SF-1
(steroidogenic factor-1). We have now purified and sequenced
this protein from immature porcine testes, identifying it as the
nuclear phosphoprotein SET; a role for SET in transcription was not
established previously. Binding of bacterially expressed human and rat
SET to the DNA site at -418/-399 of the rat P450c17 gene
transactivates P450c17 in neuronal and in testicular Leydig cells. We
also found SET expressed in human NT2 neuronal precursor cells,
implicating a role in neurosteroidogenesis. Immunocytochemistry and
in situ hybridization in the mouse fetus show that the
ontogeny and distribution of SET in the developing nervous system are
consistent with SET being crucial for initiating P450c17 transcription.
SET’s developmental pattern of expression suggests it may participate
in the early ontogenesis of the nervous, as well as the skeletal and
hematopoietic, systems. These studies delineate an important new factor
in the transcriptional regulation of P450c17 and consequently, in the
production of DHEA and sex steroids. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-aminobutyric acid
(GABAA) receptors to exert anesthetic and
anxiolytic activities (7, 8). Other neurosteroids, such as pregnenolone
sulfate, dehydroepiandrosterone (DHEA), and
DHEA-sulfate (DHEAS), act as antagonists of the
GABAA receptor (9). These same steroids may have
agonistic effects on N-methyl-D-aspartate
(NMDA) receptors (10, 11, 12, 13, 14) and may have additional effects on type 1 
receptors whose endogenous ligand(s) are unknown, but which are known
to bind haloperidol (15). Thus, some neurosteroids affect GABA- and
NMDA-associated behaviors such as anxiety, learning, and memory
(16, 17, 18, 19). We have recently (14) shown that DHEA
specifically promoted axonal growth while DHEAS specifically promoted
dendritic growth in fetal rodent neurons, and that DHEA
increased the morphological indices of synaptic contacts within
neocortical neurons (14). Some of the effects of DHEA, but
not of DHEAS, were mediated via NMDA receptor activation (14). Thus,
the neuronal regulation of DHEA synthesis is of
substantial interest in understanding fetal brain development.
The key enzyme in the biosynthesis of DHEA is P450c17, a
microsomal enzyme that has both 17
-hydroxylase and 17,20-lyase
activities, and hence converts pregnenolone to DHEA (20, 21). P450c17 expression is developmentally and regionally regulated in
the nervous system (5) by factors other than those found in the adrenal
and gonad (22, 23, 24). Expression and transcriptional regulation of the
gene for P450c17 [formally termed CYP17 (25)] are regulated by ACTH
in the adrenals and LH in the gonads via the cAMP/protein kinase A
signaling pathways, but is not mediated through binding of CREB (cAMP
response element binding protein) to a consensus cAMP-
responsive DNA element (22, 26, 27, 28).
We previously identified a region of the rat P450c17 gene between -447/-399 bp upstream from the transcriptional start that is important for both basal and cAMP-regulated transcriptional activities in the adrenal and testis (29). This region was regulated by several members of the orphan nuclear receptor family, including SF-1 (steroidogenic factor-1), NGF-IB (nerve growth factor inducible protein B), and COUP-TF (chicken ovalbumin upstream promoter transcription factor). We also identified the binding sites for two additional nuclear proteins operationally termed steroidogenic factor inducer of transcription-1 and -2 (StF-IT-1 and StF-IT-2) that are important for rat P450c17 gene transcription (29). We have now demonstrated that the StF-IT-1 site is also transcriptionally active in human neuronal precursor NT2 cells, purified StF-IT-1, and identified it as the product of the protooncogene SET, a protein not previously implicated in transcriptional regulation. Our present results now show that this novel transcription factor plays a role in the regulation of P450c17 in the developing nervous system.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Two additional
factors, which we called StF-IT-1 and StF-IT-2, also bound to this same
region (29). StF-IT-1 bound specifically to the -418/-399 region
(Fig. 1B) and transcriptionally activated basal activity from this
region of the rat P450c17 gene in both Leydig (Fig. 1C, left
graph) and adrenocortical cells, when this region was ligated to a
heterologous promoter, strongly suggesting that StF-IT-1 is
important for endogenous P450c17 transcription. This same -418/-399
region could also be transcriptionally activated when transfected into
human neuronal precursor NT2 cells (Fig. 1C, right graph).
This region was only minimally affected by cAMP in MA-10 and NT2 cells
and is therefore considered to be a strong basal element, rather than a
cAMP-regulated element. StF-IT-1 appeared to be a novel DNA binding
protein that shared no similarity to other known transcriptional
factors in its DNA sequence specificity for binding or activity,
although the binding sequence of StF-IT-1 contained a variant (AGGAGA)
of the estrogen receptor half-site sequence AGGTCA (29).
|
), while the majority of the
protein was retained on the column. An additional protein binding to
the -418/-399 rP450c17 oligonucleotide was eluted from the column in
fractions 24–26. This protein-DNA complex was slightly larger than the
StF-IT-1-DNA complex seen in the starting material and was not analyzed
further. Some StF-IT-1 was also retained on the Mono-S column and was
eluted in fractions 26–29. However, this retention of StF-It-1 was
only seen when a large amount of protein (>20 mg) was applied to the
Mono-S column.
|
).
The N-terminal sequence of 20 amino acids from the 
16-kDa protein
identified it as streptavidin that came from the affinity column. The
N-terminal sequence of 60 amino acids from the 24-kDa protein shared
95% identity with the human SET protein (GenBank accession no.
Q01105), protein phosphatase inhibitor 2A (I2PP2A) (GenBank accession
no. U60823), and PHAP II (HLA-DR associated protein II, EMBL accession
no. X75091) (Fig. 2C). The high degree of amino acid sequence identity
among porcine StF-IT-1, human SET, human I2PP2A, and human PHAP II
suggests that all three are closely related and possibly identical.
However, it is not certain that StF-IT-1 is identical to SET, or simply
encoded by a very similar gene. The absence of the first 16 amino acids
of human SET in our purified porcine protein may reflect proteolytic
processing, alternate splicing, protein degradation during the
purification, a species-specific difference, or the possibility that
SET and StF-IT-1 are encoded by highly homologous, but not identical,
genes.
SET Is a Site-Specific DNA Binding Protein
The identification of human SET and its name derive from the
description of a chromosomal translocation found in a patient with
acute undifferentiated leukemia (patient "SE
translocation") (30). In this patient, the gene set
(located on chromosome 9q34, centromeric of c-abl) was fused
to the gene can, also found on chromosome 9q34. SET is an
acidic nuclear phosphoprotein (pI 3.9) of 277 amino acids, with a total
of 33% acidic amino acids at the carboxy terminus (31, 32). SET
was previously purified from several sources, and its expression was
associated with roles other than transcription, including its role as a
phosphatase 2 inhibitor and activator of adenovirus replication (30, 33, 34, 35, 36).
To determine whether the StF-IT-1 protein binding to -418/-399 was
immunologically related to SET, we obtained antisera to human SET
peptides (generously provided by Dr. Terry Copeland, National Cancer
Institute). Using the -418/-399 rat P450c17 oligonucleotide as probe
and mouse testicular Leydig MA-10 cells or adrenocortical Y-1 cells as
a source of nuclear extract, addition of anti-SET antisera blocked the
formation of the complex whereas preimmune serum did not (Fig. 3A). Similar results were obtained with
nuclear proteins prepared from mouse adrenocortical ST-R cells (37) and
rat C6 glioma cells (data not shown). When the SET-related protein(s)
found in the nuclear extracts were depleted by immunoprecipitation with
SET antiserum, the immunodepleted extract did not form the StF-IT-1-DNA
complex with the -418/-399 rat P450c17 oligonucleotide (Fig. 3B).
Thus, the StF-IT-1-DNA complex seen in vivo with the
-418/-399 region of the rat P450c17 gene is due to binding of a
protein immunologically related to SET.
|
, left) as
efficiently as the StF-IT-1 protein from MA-10 cells (Fig. 3C, right); furthermore, wild-type, but not mutant
-418/-399 oligonucleotides would compete for human SET binding (Fig. 3C, left). This binding pattern of human SET could be
duplicated by incubating -418/-399 with MA-10 cell extract (Fig. 3C, right). Thus, human SET binds to -418/-399
indistinguishably from MA-10 cell StF-IT-1.
To determine whether authentic SET protein exerts the same biological
effects as those we have previously shown for StF-IT-1 (29), we
assessed the ability of rat SET to transactivate the rat -418/-399
sequence in human NT2 neuronal precursor cells (Fig. 4A). Rat SET cDNA, cloned by RT-PCR, was
inserted into a eukaryotic expression vector, and NT2 cells were
cotransfected with this vector and with a luciferase reporter gene
under the control of the -418/-399 oligonucleotide linked to the
32-bp minimal promoter of the herpes simplex virus thymidine kinase
gene (-418/-399TK32LUC). Because NT2 cells express low levels of SET
endogenously, the wild-type -418/-399TK32LUC vector shows luciferase
activity in the absence of the SET expression vector. A mutant
-418/-399TK32LUC vector in which the variant of the estrogen
receptor half -site was changed, however, shows no more activity than
the TK32LUC construct alone, consistent with endogenous NT2 SET acting
through the wild-type rat -418/-399 sequence. When the cells are
cotransfected with the rat SET expression vector, activity from
wild-type -418/-399TK32LUC increased 550% above the level without
the rat SET vector, but the mutant -418/-399TK32LUC still had no more
activity than the TK32LUC control. Thus SET specifically binds to the
TCTCCTCAA sequence of the rat P450c17 promoter to elicit a profound
increase in basal transcription.
|
). We previously
made a deletional construct containing 476 bp of the 5'-flanking region
of the rat P450c17 DNA that contained the -418/-399 region, ligated
to the reporter gene 
luciferase (-476 -Luc), and showed that
this construct was transcriptionally active in mouse Leydig MA-10
and adrenocortical Y-1 cells (22). This vector was transfected
into N2A cells, in the absence or presence of the SET expression
vector. The -476 -Luc vector has some activity in the absence of
the SET expression vector, but it is less than the -418/-399TK32LUC
vector because it contains a binding site for the transcriptional
inhibitor COUP-TF (29). Nevertheless, when the SET expression vector
was cotransfected into the N2A cells, activity from the -476 -Luc
vector increased 10-fold, indicating that SET could increase
transcription from the rat P450c17 gene in its natural DNA context.
Developmental Analysis of SET Expression in the Nervous System
We have previously shown that P450c17 is expressed in specific
regions of the developing fetal rodent brain, even though SF-1, which
appears to be required for adrenal and gonadal expression of P450c17,
is not expressed in the fetal brain where P450c17 is expressed (5, 24).
To determine whether SET is expressed in the same brain regions that
express P450c17, we used immunocytochemistry to colocalize SET and
P450c17 protein, and in situ hybridization to colocalize SET
mRNA.
Early in brain development, SET mRNA was expressed from E10.5 in the
prosencephalon (Fig. 5, panels A, C, and
D), the structure of the rhombencephalon from E11.5 [Fig. 5D, mesencephalon (Mes) and metencephalon (Met)] but not in the developing
rhombencephalon at E10.5 (which is negative in Fig. 5B), the
prosencephalon, rhombencephalon (mesencephalon and metencephalon) and
diencephalon at E13.5 (Fig. 5E), and in the basal diencephalon,
pituitary, and hindbrain (shown at E18.5 in Fig. 6K). These data suggest that SET is
expressed in the brain in an antero-posterior gradient. At this
same time in development (E11), SET protein was coexpressed with
P450c17 in the developing neural tube in the lateral motor column
(data not shown). In addition, SET mRNA expression was restricted to
the dorsal and ventral segments of the neural tube. SET protein was
also expressed in the notochord at E11 although P450c17 protein was not
expressed there (data not shown).
|
|
). SET mRNA and P450c17
protein colocalized within the subplate at this time (data not
shown). At this same time in development, SET was widely expressed in
the embryo (Fig. 6). Sites of expression included the cochlea,
paraspinal muscle, thymus, whiskers, skin, intestine, spinal cord,
dorsal root ganglia, cartilage, and trigeminal ganglia (Fig. 6). SET
mRNA was also expressed in sites related to hematopoiesis and
development of the vasculature, such as the developing liver, the
branchial arches (Fig. 5A, ba), and the walls of vessels (spinal artery
and vessels in the adult ovary, not shown). SET protein and mRNA
expression were developmentally regulated and were largely reduced
around birth. Nevertheless, at P9, SET mRNA could still be detected in
the hippocampus, cortex, thalamus (Fig. 5G), hypothalamus, septum, and
in both the granular layer and Purkinje cells of the cerebellum (Fig. 6K), but its expression disappeared in the adult.
Both SET and P450c17 were expressed in structures derived from the
migration of neural crest cells. P450c17 protein was found mainly in
structures derived from the cranial neural crest (5) while SET mRNA was
found in structures derived from the cranial neural crest
(cranial-facial bones and cartilage such as cochlea, Fig. 6A; smooth
muscles of the back, Fig. 6B; and thymic cells, Fig. 6C) as well as
from the trunk neural crest (dorsal root ganglia, Fig. 6H; skin, Fig. 6E) and cardiac neural crest (walls of the large arteries,
e.g. spinal artery, data not shown). SET mRNA was also
expressed in mesodermally derived structures such as the sclerotomes
(E10.5 and E11.5, Fig. 5D) and in the vertebral cartilage (Fig. 6H).
SET mRNA was expressed in developing bone, including primordial
cartilage and limb bud, and later in the skeleton in the ribs, femur,
skull, and jaws.
SET mRNA and protein were found in the same neurons that express
P450c17: in the peri-locus c
ruleus nucleus (not shown), trigeminal
ganglia (Fig. 7, A and B), in the pontine
nucleus (Fig. 7, C and D) as well as in the cortical subplate (E18.5,
Fig. 5F). SET mRNA was also detected in the cortex (Fig. 5F) and in
mesencephalic, hypothalamic, thalamic, and septal nuclei (not shown)
where P450c17 was not expressed, suggesting that SET may regulate other
genes in those structures. However, P450c17 was never expressed in
regions that did not express SET.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and SETß, that differed at their 5'-ends. SET
cDNA has an open reading frame of 867 nucleotides, while SETß cDNA
has an open reading frame of 831 nucleotides. SET cDNA encodes a
protein that has 36 amino acids at its amino terminus that differ from
the first 24 amino acids of the protein encoded by SETß cDNA; the
remaining cDNA sequences are identical between the two SET species.
These differences may arise from alternative splicing or may represent
sequences from two different genes. SET mRNA was much less abundant
than SETß mRNA in all tissues examined, including rat brain, heart,
lung, and kidney and thus may represent only a minor transcript. The
sequence of our porcine StF-IT-1 corresponds with the sequences found
in SETß, as the first eight amino acids of our porcine protein
correspond to those predicted for amino acids 17–24 of SETß and not
of SET. SET is a potent and specific noncompetitive inhibitor of protein phosphatase 2A (PP2A) (39), a protein involved in the regulation of normal cell growth (40, 41, 42). SET may also be involved in the alteration of chromatin structure to promote increased gene transcription. The human homolog, HRX, of the Drosophila trithorax protein interacts with SET and protein phosphatase 2A (43). Experiments have suggested that HRX and associated proteins may affect nucleosome assembly, alter chromatin structure, and hence alter access of transcription factors to DNA. SET also has amino acid sequence similarity to the Drosophila nucleosome assembly protein NAP-1 (44, 45), a core histone shuttle that delivers histones H2A and H2B from the cytoplasm to the chromatin-assembly machinery in the nucleus in a cell cycle-dependent manner. Yeast NAP-1 can stimulate binding of transcription factors by a mechanism involving nucleosome displacement (46). A 43-kDa Xenopus homolog of SET, as well as a 60-kDa Xenopus NAP-1 protein, both interact specifically with B-type cyclins (47). Hence, in Xenopus, NAP/SET proteins may regulate cell cycle. Finally, a 39- or 41-kDa protein purified from HeLa cells, called template activating factor-1, or TAF-1 (48) was shown to be identical to SET and was shown to stimulate adenovirus core DNA replication (35, 36). Thus in this context, SET and SET-like proteins may be involved in chromatin remodeling and direct transcriptional activation, perhaps inappropriately, leading to leukemogenesis.
Our data now demonstrate that, in addition to the other functions attributed to SET or SET-like proteins, SET is a DNA binding protein and transcriptional activator, that plays a role in transcription of the gene for P450c17, and possibly other genes, as we find SET expressed in cells not expressing P450c17. The structure of SET protein resembles no known class of transcription factor identified so far. It is a protein of 277 amino acids, with a long acidic tail of 53 amino acids at its carboxy terminus. This acidic region of TAF-1, which is identical to SETß, is essential for stimulation of replication from adenovirus DNA and for interaction with cellular histones (48, 49, 50). In addition, the replication activity of TAF-1 is dependent upon dimerization (51), and it is presumed that the acidic tail plays a role in this dimerization. It is unknown whether these regions may play roles in transcriptional activation as well.
The sites of SET expression suggest that P450c17 and other target genes may be involved in neural induction and more generally in cell differentiation. Thus the set-can gene fusion that initially identified the set gene may cause undifferentiated leukemia by creating a SET-CAN fusion protein that alters SET protein function, rather than by altering CAN protein function. Removal of the seven carboxy-terminal amino acids by fusion with CAN may alter the transactivating function of SET. By reducing SET activity, cells may not enter the last stages of differentiation and may be propagated as abnormal pluripotent stem cells that result in undifferentiated leukemia.
In another type of leukemia, acute nonlymphocytic leukemia, the can gene is fused to the dek gene, resulting in leukemia-specific, chimeric dek-can mRNA and fusion protein. Recent experiments have shown that DEK also binds to specific DNA sequences to increase gene transcription (52). The DNA binding site for DEK is not similar to the binding site we identified for SET. Also, it has not been demonstrated whether DEK increases or decreases gene transcription. Thus, the genetic mechanism of leukemogenesis may be similar for both set-can and dek-can translocations, but the downstream genes affected by these translocations are likely to be different. With the identification of SET as a DNA binding protein, it may now be possible to search for other SET-target genes. Searching the GenBank database has not yet produced other genes.
Recent experiments have proposed a novel mechanism for leukemogenesis in chromosomal translocation-generated oncoproteins (53). Both DEK-CAN and SET-CAN encode nuclear fusion proteins, called nucleoporins, that contain a Phe-Gly (FG) repeat region within CAN protein. This repeat region has been shown to interact with the transcriptional coactivator CREB binding protein (CBP) and p300. As a result, transcriptional activators fused to proteins such as CAN show increased transcriptional activity. As we have now shown that SET is a potent transcriptional activator, the fusion of the set to the can gene in patients with acute undifferentiated leukemia may result in a SET-CAN fusion protein that is able to recruit additional coactivators to increase the transactivation function of SET further, thus leading to oncogenesis.
Role of SET in Nervous System Development
The pattern of expression of SET in the developing neural tube
supports the hypothesis that SET may activate genes involved in the
organogenesis of the spinal cord. SET is expressed both in the
developing neural tube in restricted dorsal and ventral areas as well
as in the notochord, suggesting that its expression may be involved in
turning on genes involved in regulating neuronal induction in these
regions. One of these genes is P450c17, whose expression lags behind
SET in the developing motor neurons. From the sites of SET expression
and from roles determined for SET homologs, we believe that SET
expression in regions where P450c17 is expressed, and in other regions
where P450c17 is not expressed, may be related to determination of cell
fate. For example, SET is expressed in cells derived from the migrating
neural crest and is found in cranio-facial cartilage. SET is also
present from the early stages of the determination of somites to the
formation of bones.
SET expression is developmentally regulated, and its expression declines in parallel with the decline in P450c17 expression in the central nervous system. COUP-TF is a nuclear factor that competes for the SET binding site in the rat P450c17 gene, and hence is a potent repressor of P450c17 (29). COUP-TF, unlike SET, is not expressed early in development, but rather is expressed later in embryogenesis in regions that express P450c17 (54). We postulate that increased expression of COUP-TF inhibits P450c17 expression in the central nervous system. Thus SET activation of P450c17 transcription in the developing nervous system may ultimately result in increased DHEA production, which may be an important signal for modulating neurotransmission to trigger formation of neuronal circuits.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-32P]ATP (50 µCi, 3000 mCi/mmol) to
monitor the presence of the oligonucleotide. Then, the oligonucleotides
were blunt-ended by Klenow DNA polymerase using dTTP, dGTP, and dCTP
and biotin-dATP (final concentration of 20 µM each
nucleotide in a volume of 100 µl). The biotin and
32P double-labeled oligonucleotides were purified
by chromatography on NAP-10 columns (Pharmacia Biotech,
Piscataway, NJ) and used for the protein binding reactions.
Preparation of Homogenates of Immature Porcine Testes
Immature porcine testes were collected from a commercial pig
farm and shipped to the laboratory on ice. The tissue was dissected and
homogenized with a Dounce homogenizer in buffer containing 60
mM KCl, 15 mM NaCl, 15 mM HEPES, pH
7.8, 14 mM mercaptoethanol, 0.3 M sucrose, and
a cocktail of protease inhibitors (0.5 mM
phenylmethylsulfonylfluoride, 0.5 µg/ml pepstatin, 0.5 µg/ml
antipain, and 0.5 µg/ml leupeptin). The crude tissue homogenate was
centrifuged at 3000 x g for 5 min to remove the large
tissue debris, and the supernatant was collected and stored in aliquots
at -70 C. We also made a nuclear preparation as described
previously(55), tested a number of known nuclear proteins, and found
that all these known nuclear proteins, including StF-IT-1, StF-IT-2,
and SF-1, were present in the cytoplasmic fraction (supernatant).
Therefore, we used the supernatant for further purification.
Protein Purification by Chromatography
The porcine testicular extract was dialyzed against
equilibration buffer (MES 20 mM, pH 5.5) (MES is
2-[N-morpholino]ethanesulfonic acid), centrifuged for
1 h at 100,000 x g, and the supernatant was
applied to a 10 ml Protein Pack SP 15 HR FPLC column (Waters Corp., Milford, MA), previously equilibrated with the same
buffer. Proteins were detected by UV absorption at 280 nm using a
Waters UV Detector, model 440. Proteins not retained on the column were
eluted in MES buffer until the OD280 returned to
0. Proteins retained on the column were eluted with a linear gradient
of NaCl (0–0.5 M NaCl) generated by a Waters
Controller System, model 600E, at a flow rate of 1 ml/min. About one
third of the protein was found in the flowthrough of the column, and
two thirds of the protein was retained.
Fractions containing DNA binding activity (flowthrough) were pooled and dialyzed overnight against the DNA binding buffer containing 20 mM HEPES, pH 7.9, 50 mM KCl, 4 mM Tris-HCl, pH 7.9, 5 mM EDTA, pH 7.9, 1 mM dithiothreitol, and 1 mM phenylmethylsulfonylfluoride. The dialysate was centrifuged at 13,000 x g for 5 min to remove any precipitate before being used for DNA binding assays. The biotinylated oligonucleotide (20 µg) was added to the sample together with nonspecific calf thymus DNA that was previously sonicated, heated at 100 C for 5 min, and used at final concentration of 150 µg/ml. The DNA binding reaction was performed at 4 C for 2–3 h before adding the streptavidin-agarose (1 ml prewashed twice with 1 M KCl in 1x binding buffer, and then five times with 1x DNA binding buffer to remove the salt). After incubation with streptavidin at room temperature for 1 h, the sample was centrifuged at 3,000 x g for 10 min to pellet the streptavidin-biotinylated oligonucleotide conjugates. The supernatant was carefully removed by pipette, and the conjugated agarose matrix was transferred to an Eppendorf tube to remove unconjugated protein. The conjugated agarose matrix was washed ten times with 1x binding buffer (1 ml each time, mixed manually for 1 min), removing the supernatant each time. The final elution of protein specifically bound to the oligonucleotide was performed by adding 400 µl elution buffer twice (1 M KCl in 1x binding buffer), incubating the sample with the elution buffer for 5 min at room temperature, and separating from the agarose matrix by centrifugation at 3000 x g. The elute was stored at -20 C until further analysis.
Protein Sequencing
The final 1 M KCl eluate from the oligonucleotide
affinity column was dialyzed against ammonium bicarbonate (50
mM, pH 8.3), dried under vacuum, resuspended in SDS sample
buffer, and separated by 10% SDS-PAGE. Immediately after
electrophoresis, proteins were transferred to a PVDF membrane
(Bio-Rad Laboratories, Inc. Richmond, CA) in 10
mM 3-[cyclohexylamino]-1-propanesulfonic acid (CHAPS)
buffer, pH 11, in 10% methanol. Transferred proteins were stained in
0.1% Coomassie Brilliant Blue R-250 in 50% methanol/1% acetic acid,
and were destained in 50% methanol. The band of interest was excised
from the PVDF membrane and subjected to N-terminal microsequencing on a
vapor phase Beckman-Porton PI 2090 sequencer (Beckman Coulter, Inc., Fullerton, CA), using the Edman degradation procedure. The
Edman degradation cycles had yields of more than 90% and contained
about 45–50 pmol of material per cycle. Sequences obtained were
searched for homology with sequences in the SWISS-PROT database, using
the FASTA search of the GCG program.
Gel Mobility Shift Assays
Gel mobility shift assays were performed as described previously
(22, 23, 29). Whole-cell extracts from MA-10 and N2A cells were
prepared according to previously published procedures (22, 56). A
wild-type oligonucleotide was derived from sequences 399 to 418 bp
upstream from the transcription initiation site of the rat P450c17 gene
(called "-418/-399") (29). A number of mutant oligonucleotides
were also used as the unlabeled competitors for the wild-type probe
(Table 1). Oligonucleotide probes were end labeled
using [-32P] ATP and T4 polynucleotide
kinase and mixed with 10 µg of the proteins in the presence of 100
µg/ml poly dI/dC, 50 µg/ml salmon sperm DNA, 5 mM
dithiothreitol, and 1 mg/ml BSA and incubated at room temperature for
40 min. One quarter of the total reaction was loaded onto a 6%
nondenaturing polyacrylamide gel, using 0.5x Tris-borate-EDTA
as a running buffer. The dried gel was then exposed to x-ray film. In
the case of monitoring the DNA binding activity from the column
fractions, we used 10 µl of each column fraction in the gel shift
assay reaction. The film exposure time varied from 2 h to
overnight depending upon the protein concentration and the DNA binding
affinity of the specific binding protein.
|
Full-length rat SET cDNA was also cloned into the eukaryotic expression vector pCR3 (Invitrogen, San Diego, CA), amplifying SET RNA from rat kidney RNA, using rat oligonucleotide sequences as primers (5'-primer: 5'-ATGTCTGCGCCGACGGCC-3'; 3'-primer: 5'-CTAGTCATCCTCGCCTTCATCCTC-3') (38).
Analysis of SET and P450c17 mRNAs and Proteins
A 455-bp rat SET cDNA fragment [nucleotides (nt) 227–682]
(38), prepared by RT-PCR amplification of rat kidney RNA using
rat-specific oligonucleotides, was cloned into pKS
(Stratagene, La Jolla, CA) and generated a 529-nt probe. A
120-bp EcoRI-BamHI rat P450c17 cDNA fragment
cloned into pKS generated a 171-nt probe (57). In situ
hybridization of SET mRNA was performed on fresh frozen embryos (4, 5)
or on paraffin sections obtained commercially (Novagen), using
35S-labeled RNA probes. Immunocytochemistry on
fresh frozen tissues was performed as described previously (4, 5) using
antibodies against human SET peptides (31) and recombinant human
P450c17 (58), and using a fluorescein isothiocyanate-conjugated second
antibody. P450c17 antibodies were used at a 1:2000 dilution for
immunocytochemistry. The SET antibodies were generated against three
human SET peptides: SP-1, amino acids 3–16; SP-2, amino acids 44–56;
and SP-3, amino acids 169–181. Antibody SP-2 worked best in gel shift
assays while antibody SP-3 worked best for immunocytochemistry and was
used at a 1:3000 dilution.
Construction of the Rat P450c17 Oligonucleotide-TK-LUC Expression
Plasmids
Rat P450c17 oligonucleotides were cloned into a luciferase
expression vector with a minimal promoter from the TK gene of herpes
simplex virus (TK32LUC) as described (22). 5'-Deletional constructs of
the rat P450c17 gene, ligated to the reporter gene -luciferase, were
described previously (22). All constructs were confirmed by DNA
sequencing to determine oligonucleotide copy number, orientation, and
sequence. Plasmids containing only a single copy of the oligonucleotide
cloned in the 5' 
3' direction were used for transfection
experiments.
Cell Culture, Transfections, and Luciferase Assays
Mouse Leydig MA-10 cells (59) were grown as described previously
(22). Human neuronal NT2 precursor cells (Stratagene, La
Jolla, CA) were cultured in 50% Ham’s F12/50% DME H21, 10% FBS, 1%
glutamine, 1% penicillin/streptomycin. Plasmid DNAs were transfected
into MA-10 or NT2 cells by lipofection, using the Fugene 6 transfection
reagent (Roche Molecular Chemicals, Indianapolis, IN).
When vectors expressing SET were cotransfected with reporter luciferase
constructs, the molar ratio of these two plasmids was 1:1. DNA
concentrations were equalized by the addition of the cloning vector
pCR3. Cells stimulated with cAMP were treated with 1 mM
8-Br-cAMP for the indicated times. Luciferase assays and data analysis
were as described elsewhere (60), using a Monolight 1500 luminometer
(Analytical Luminescence Laboratory, San Diego, CA) and a
luciferase assay system (Promega Corp., Madison, WI).
Cellular protein concentrations were assayed using the BCA protein
assay kit (Pierce Chemical Co.).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
This work was funded by NIH Grants HD-27970 (to S.H.M.) and HD-11979 (to the Reproductive Endocrinology Center, UCSF) and by a grant from the Alzheimer’s Association (to S.H.M).
1 Both authors contributed equally and should be considered co-equal
first authors. 
Received for publication November 16, 1999. Revision received February 18, 2000. Accepted for publication February 24, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-hydroxylase/17, 20
lyase): similarity to the gene for P450c21. DNA 6:437–448
-hydroxylase and 17,20-lyase activities of P450c17: contributions of
serine106 and of P450 reductase. Endocrinology 132:2498–2506This article has been cited by other articles:
 |
|
 |
|
  S.-B. Lee, D. S.-C. Ou, C.-F. Lee, and L.-J. Juan Gene-specific Transcriptional Activation Mediated by the p150 Subunit of the Chromatin Assembly Factor 1 J. Biol. Chem., May 22, 2009; 284(21): 14040 - 14049. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Shi, M. D. Schonemann, and S. H. Mellon Regulation of P450c17 Expression in the Early Embryo Depends on GATA Factors Endocrinology, February 1, 2009; 150(2): 946 - 956. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J. Kuhl, S. M. Ross, and K. W. Gaido CCAAT/Enhancer Binding Protein {beta}, But Not Steroidogenic Factor-1, Modulates the Phthalate-Induced Dysregulation of Rat Fetal Testicular Steroidogenesis Endocrinology, December 1, 2007; 148(12): 5851 - 5864. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Ozbek, A. Kandilci, S. van Baal, J. Bonten, K. Boyd, P. Franken, R. Fodde, and G. C. Grosveld SET-CAN, the Product of the t(9;9) in Acute Undifferentiated Leukemia, Causes Expansion of Early Hematopoietic Progenitors and Hyperproliferation of Stomach Mucosa in Transgenic Mice Am. J. Pathol., August 1, 2007; 171(2): 654 - 666. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. Mellon, S. R. Bair, C. Depoix, J.-L. Vigne, N. B. Hecht, and P. B. Brake Translin Coactivates Steroidogenic Factor-1-Stimulated Transcription Mol. Endocrinol., January 1, 2007; 21(1): 89 - 105. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. V. Pandey, S. H. Mellon, and W. L. Miller Protein Phosphatase 2A and Phosphoprotein SET Regulate Androgen Production by P450c17 J. Biol. Chem., January 24, 2003; 278(5): 2837 - 2844. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Qu, Q. Li, H.-Y. Lim, N. S. Cheung, R. Li, J. H. Wang, and R. Z. Qi The Protein SET Binds the Neuronal Cdk5 Activator p35nck5a and Modulates Cdk5/p35nck5a Activity J. Biol. Chem., February 22, 2002; 277(9): 7324 - 7332. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Lin, J. W. M. Martens, and W. L. Miller NF-1C, Sp1, and Sp3 Are Essential for Transcription of the Human Gene for P450c17 (Steroid 17{alpha}-hydroxylase/17,20 lyase) in Human Adrenal NCI-H295A Cells Mol. Endocrinol., August 1, 2001; 15(8): 1277 - 1293. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
