(Received for publication, March 13, 1997, and in revised form, March 31, 1997)
From the Department of Molecular Medicine, Institute of Biotechnology, University of Texas Health Science Center, San Antonio, Texas 78245-3207
Nicotinic acetylcholine receptors constitute a
multigene family (2-
9,
2-
4) expressed in discrete
temporal and spatial patterns within the nervous system. The receptors
are critical for proper signal transmission between neurons and their
targets. The molecular mechanisms underlying receptor gene expression
have not been completely elucidated but clearly involve regulation at
the level of transcription. We previously identified a novel 19-base
pair (bp) transcriptional regulatory element in the promoter region of
the rat
4 subunit gene. This 19-bp element interacts specifically
with DNA-binding proteins enriched in nuclear extracts prepared from
adult rat brain. Using a combination of cellulose-phosphate, DNA-cellulose, and DNA sequence-specific affinity chromatographies, we
purified the 19-bp element binding activity approximately 19,000-fold. Analysis by denaturing gel electrophoresis revealed the presence of
four polypeptides in the most purified fraction, ranging in molecular
masses between 31 and 114 kDa. Peptide sequence analysis revealed that
one of the polypeptides is the bovine homologue of the transcriptional
regulatory factor, Pur
. Electrophoretic mobility shift assays
indicated that Pur
interacts directly and specifically with the
19-bp element. In addition, mobility shift assays using an anti-Pur
monoclonal antibody revealed the presence of Pur
, or an
immunologically related protein, in nuclear extracts prepared from
brain tissue. We hypothesize that the interaction between Pur
and
the 19-bp element is critical for proper expression of the
4 subunit
gene.
One of the key events that takes place during development of the
nervous system is the formation of synapses. Numerous lines of evidence
indicate that expression of neurotransmitter sensitivity is central to
synaptogenesis (reviewed in Ref. 1). It is also clear that changes in
sensitivity to neurotransmitters are due to changes in the expression
of neurotransmitter receptors. Recent advances toward a molecular
understanding of the formation of nicotinic cholinergic synapses in the
nervous system were made with the isolation of a family of genes
(2-
9 and
2-
4) encoding subunits of the nicotinic
receptors for acetylcholine (nACh1;
reviewed in Ref. 2). Functionally diverse nACh receptors can be
generated by distinct combinations of the subunits in vitro (3-5). In addition, each of the subunits exhibits discrete temporally and spatially restricted expression patterns in vivo (2).
Despite these important advances, however, the cellular and molecular mechanisms controlling the expression of the nACh receptor genes are
relatively obscure, although several recent studies suggest that both
positive and negative transcriptional regulatory mechanisms are
involved in receptor gene expression (6-16).
Three of the nACh receptor genes, those encoding the 4,
3, and
5 subunits, are tightly linked within the rat genome, spanning only
approximately 60 kilobase pairs (17). This genomic organization, coupled with the high nucleotide sequence similarities between the
three genes, suggests that the genes arose through tandem duplication
of a common ancestral gene as has been suggested previously (17). On
the other hand, the genomic organization may also reflect a regulatory
mechanism that ensures co-expression of the genes in the appropriate
cell types at the appropriate developmental time. Such a mechanism
would be consistent with recent data indicating the presence, in
vivo, of receptors consisting of
4,
3, and
5 subunits
(18). However, the expression of these genes is not completely
overlapping (2) and, thus, some other mechanism must account for this
disparity. One possibility is that while each of the subunit genes is
independently regulated transcriptionally, they share some common
regulatory features, which are active in the appropriate developmental
and cellular environments. Several laboratories, including ours, have
characterized distinct promoter regions for the
4 (10, 11, 14) and
3 (9, 15, 16) subunit genes. Within these regions, several
transcriptional regulatory elements have been identified. Thus far, the
only regulatory element common to both the
4 and
3 genes is a
consensus Sp1-binding site, which appears critical for transcription of
the two genes (9, 14). Our analysis of the rat
4 subunit promoter
region led to the identification of three elements, E1-E3, which may play roles in
4 gene expression (14). E2 is the aforementioned Sp1-binding site, E3 shares some sequence homology with the consensus AP1-binding site, and E1 is a novel 19-bp regulatory element (14). E1,
characterized by three CCCT repeats, forms specific complexes with
proteins present in nuclear extracts prepared from neuronal cells (11,
14). We have now used E1 as an affinity ligand in a purification scheme
to isolate, from bovine brain nuclear extracts, the nuclear proteins
which bind specifically to this element. In the most highly purified
fraction, four polypeptides with apparent molecular masses of 31, 43, 65, and 114 kDa were detected. We termed these E1-binding proteins,
neuronal nACh receptor promoter-binding proteins (NARP31, -43, -65, and -114).
Peptide sequence analysis indicated that NARP43 is the bovine homologue of Pur
, a DNA-binding protein previously shown to stimulate
transcription of the myelin basic protein gene in oligodendrocytic
cells (19). Electrophoretic mobility shift assays (EMSA) demonstrated
that Pur
is present in nuclear extracts prepared from brain and
binds directly and specifically to E1, raising the possibility that this interaction plays a role in the expression of the
4 subunit gene.
Buffer C consisted of 25 mM HEPES (pH 7.8), 50 mM KCl, 0.1 mM EDTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride, and 20% glycerol. Buffer D consisted of 50 mM Tris (pH 7.5), 50 mM NaCl, 1 mM EDTA, 0.5 mM DTT, 1 mM phenylmethylsulfonyl fluoride, and 10% glycerol. Buffer E consisted of 20 mM HEPES (pH 7.5), 50 mM NaCl, 5 mM EDTA, 1 mM DTT, 10% glycerol. Binding buffer consisted of 20 mM HEPES (pH 7.4), 5 mM DTT, 1 mM MgCl2, 50-100 mM KCl, and 6% glycerol. Buffer Z consisted of 25 mM HEPES (pH 7.6), 10 mM MgCl2, 1 mM DTT, 20% glycerol, and 0.1% Nonidet P-40.
Substrate Preparation for Electrophoretic Mobility Shift AssaysThe sequence of E1 used as a probe and competitor in EMSA
is shown in Fig. 1 along with the sequence of a mutated form of the
element used as a competitor in EMSA. The E1 complementary oligonucleotides (from Cruachem, Inc.) were annealed and then radioactively labeled with [-32P]ATP (DuPont NEN)
using T4 polynucleotide kinase (Promega). The specific activities of
the probes were typically between 14,000 and 17,000 cpm/fmol.
Competitor oligonucleotides were prepared by annealing equal amounts of
non-radioactive complementary oligonucleotides.
Preparation of Nuclear Extracts
Nuclear extracts were prepared as described by Christy et al. (20). Bovine brains were obtained within 1 h of death from a local slaughterhouse (Kiolbassa Provision Company), while fresh rat brains were from adult female Harlan Sprague Dawley animals (Charles River Laboratories). The final rat brain nuclear extract pellet was resuspended in buffer C, while the final bovine brain nuclear extract pellet was resuspended in buffer D in preparation for phosphocellulose chromatography (see below). The nuclear extracts were dialyzed overnight in the buffer in which they were resuspended. Insoluble material was removed by centrifugation.
Protein AssaysDuring phosphocellulose and DNA-cellulose chromatography, protein was monitored by measuring absorbances at 280 nm. Protein concentrations were determined by using bicinchoninic acid (Pierce).
Electrophoretic Mobility Shift AssaysE1 binding activity
was detected using EMSA as described previously (11). To demonstrate
sequence specificity of the DNA-protein complexes, competition
experiments were done with a 15-min preincubation of unlabeled
double-stranded oligonucleotides prior to the addition of labeled
oligonucleotide. The amount of DNA present in specific DNA-protein
complexes was quantified using a Molecular Dynamics PhosphorImager and
Imagequant software. Antibody supershift experiments were performed by
incubating 1 µg of a monoclonal antibody directed against Pur
(9C12, Ref. 24) or a nonrelated preimmune serum with 0.5 µg of bovine
brain nuclear extract for 10 min at 37 °C before the addition of
specific probe. Following a 1-h incubation at 4 °C, the binding
reactions were analyzed by electrophoresis through 6% native
acrylamide gels and visualized by autoradiography.
Crude nuclear extract
from bovine brain (365 mg) was applied directly onto a 1.6 × 50-cm P11 phosphocellulose column (Whatman) equilibrated with 50 mM Tris-HCl (pH 7.5) and 50 mM NaCl. After washing with three column volumes of the same buffer, bound proteins were eluted with a 1-liter linear gradient of NaCl from 50 mM to 800 mM in 50 mM Tris (pH
7.5). Approximately 9-ml fractions were collected and assayed for E1
binding activity using EMSA as described above. Fractions with binding
activity (11 mg) were pooled, concentrated by ultrafiltration, dialyzed
against buffer E, and then loaded onto a native calf thymus
DNA-cellulose column (10 ml), which was equilibrated with buffer E. After washing with three column volumes of the same buffer, bound
proteins were eluted with a 100-ml linear gradient of NaCl from 50 mM to 1.0 M in buffer E. Fractions containing
E1 binding activity (2.5 mg) were pooled, and dialyzed against buffer Z
in preparation for DNA sequence-specific affinity chromatography.
Preparation and regeneration of the E1 affinity column was done
essentially as described by Kadonaga and Tjian (21). Two 23-base
complementary oligonucleotides containing the E1 sequence (see Fig. 1)
were synthesized and purified by polyacrylamide gel electrophoresis.
Complementary 4-base overhangs (5-GATC-3
) were added to the E1
sequences to increase the efficiency of concatenation. The
oligonucleotides were annealed, phosphorylated, concatenated, and
subsequently coupled to cyanogen bromide-activated Sepharose 4B
(Pharmacia Biotech Inc.). Dialyzed, pooled fractions from the
DNA-cellulose column were incubated for 10 min at 4 °C with 2 mg of
poly(dI-dC)·poly(dI-dC). Insoluble material was removed by
centrifugation. The poly(dI-dC)·poly(dI-dC)-protein mixture was
applied to the affinity column equilibrated with buffer Z containing
0.1 M KCl. The column was washed with buffer Z containing 0.1 M KCl, followed by elution of E1 binding activity using
a linear gradient of KCl from 0.2 M to 1.0 M
(in buffer Z). Fractions containing E1 binding activity were pooled,
dialyzed against buffer Z containing 0.1 M KCl, reabsorbed
to the regenerated affinity resin, and eluted as described above.
Protein samples were separated on 10% SDS-polyacrylamide gels as described by Laemmli (22) and stained with silver (Bio-Rad). Polypeptides were transferred to a polyvinylidene difluoride membrane in buffer containing 12.5 mM Tris (pH 8.3), 96 mM glycine and 10% methanol. The membrane was stained with 0.1% Amido Black in 10% acetic acid. After destaining, NARP43 was excised, washed with HPLC-grade H2O, dried, and subjected to amino acid microsequencing (Harvard Microsequencing Facility) as described previously (23).
Expression and Purification of GST-PurA
glutathione S-transferase-Pur fusion construct,
GST-PUR-(1-322), was generously provided by Drs. Phang-Lang Chen and
Wen-Hwa Lee (University of Texas Health Science Center at San Antonio) (24). Expression and purification of GST and the GST-Pur
fusion proteins was carried out as described (24). The GST moiety was removed
from the GST-Pur
fusion protein (0.5 mg) by cleavage with thrombin
(1,500 units). The GST moiety was removed from the reaction mixture by
the addition of glutathione-Sepharose beads followed by centrifugation.
Sample stability, the efficiency of thrombin cleavage, and the purity
of the released Pur
were analyzed by SDS-PAGE (data not shown).
We previously identified several transcriptional regulatory
elements, E1-E3 (Fig. 1), in the rat 4 subunit gene
(14). One of these elements, E2, is a binding site for the regulatory
factors, Sp1 and Sp3 (14),2 while the
identities of the proteins interacting with E1 and E3 are unknown. The
study described below focuses upon E1 and the proteins with which it
interacts.
Our earlier analyses of the 4 subunit gene
focused upon the rat gene. As there are advantages for using bovine
brain versus rodent brain as a source of nuclear extract, we
determined whether nuclear extracts from both species yielded similar
electrophoretic mobility shift patterns when incubated with E1. As
shown in Fig. 2, this is indeed the case. Incubation of
radioactively labeled E1 with brain nuclear extracts prepared from both
species resulted in the formation of three prominent DNA-protein
complexes (Fig. 2). Competition experiments using either unlabeled
oligonucleotides corresponding to the
4 E1 sequence or to the
binding sites for the general transcription factors Sp1, TFIID, or AP1
indicated that only the unlabeled
4 oligonucleotide was capable of
interfering with radioactive DNA-protein complex formation. Similar
results have been obtained from several independent preparations of
nuclear extracts from each species (data not shown). These results
indicated that bovine brain is a suitable source of nuclear extracts
from which to purify the DNA-binding factors that interact with E1. We
describe below experiments resulting in a highly purified preparation of complex A.
Purification of the E1 Binding Activity
Purification of the
binding activity was achieved by a combination of phosphocellulose,
DNA-cellulose, and DNA sequence-specific affinity chromatographies. The
E1 binding activity bound to the P-11 phosphocellulose column and
eluted from the column as a broad peak between 425 and 680 mM NaCl (Fig. 3A). In contrast, a
DNA-protein complex with higher electrophoretic mobility that was
detected in the nuclear extract did not bind to the column (Fig.
3A). Competition experiments indicated that this faster
migrating species was not DNA sequence-specific (data not shown). The
origin of this nonspecific DNA-protein complex is unknown, although it
is clearly not complexes B or C as they are sequence-specific and bind
to the column (Fig. 3A, fraction 43). In
addition, the faster migrating species was not present in all nuclear
extract preparations (data not shown). An approximately 6-fold
purification of the specific E1 binding activity was achieved on the
phosphocellulose column.
The E1 binding activity was fractionated further by native DNA-cellulose chromatography. The binding activity eluted from the column as a sharp peak between 600 and 750 mM NaCl (Fig. 3B), indicating that the protein responsible for E1 binding also binds to DNA in a sequence-independent manner. DNA-cellulose chromatography afforded an approximately 3-fold greater purification of the E1 binding activity.
The final chromatographic step was recognition site affinity chromatography on a column with covalently linked, catenated E1 oligonucleotides. The E1 binding activity bound to the affinity column and eluted at 350-500 mM KCl (Fig. 3C). Two passages of the E1 binding activity over the affinity column resulted in approximately 960-fold purification with a 40% yield in this step. A summary of the purification data is presented in Table I and indicates that, relative to the nuclear extract, the final purification was approximately 19,000-fold with a 5% recovery of activity.
|
Analysis of the protein content of the E1 binding activity by silver staining after SDS-PAGE revealed four major polypeptides with molecular masses of 31, 43, 65, and 114 kDa (Fig. 3D). We refer to these proteins as neuronal nACh receptor promoter-binding proteins, or NARP (i.e. NARP31, NARP43, etc.; Ref. 14).
NARP43 Is the Bovine Homologue of PurIn several
preparations of the E1 binding activity, NARP43 was the most abundant
species; therefore, it was the first NARP to be subjected to amino acid
sequence analysis. Following electrophoresis, NARP43 was electroblotted
onto a polyvinylidene difluoride membrane and digested with trypsin.
The cleavage products were separated by high performance liquid
chromatography, and four peptides were sequenced. The four amino acid
sequences of these peptides were identical to sequences in the
previously described Pur sequence (Fig. 4; Ref. 25).
Pur
is a sequence-specific DNA-binding protein that has recently
been implicated in cell type-specific transcriptional regulation of
myelin basic protein in oligodendrocytes (19). Its recognition element
is purine-rich (25), as is the sequence of E1 (Fig. 1).
Pur
To determine whether
Pur does in fact interact with E1, EMSA were carried out with a
GST-Pur
fusion protein (24). As shown in Fig. 5,
GST-Pur
forms two major complexes with the radiolabeled E1
oligonucleotide, the larger of which is specific in that it can be
competed away by an excess of unlabeled wild type E1 oligonucleotide but not by an excess of a mutant E1 oligonucleotide (see Fig. 1 for the
sequences). GST alone does not form any specific complexes with E1
(Fig. 5). To confirm that Pur
interacts specifically with E1, the
GST moiety was cleaved from the GST-Pur
fusion protein, the GST
moiety was removed by thrombin cleavage (see "Experimental Procedures"), and the recovered Pur
protein was used in EMSA. As
shown in Fig. 5, purified Pur
interacts specifically with E1 forming
one major complex that can be competed by an excess of wild type E1
oligonucleotide but not by mutant E1 oligonucleotide.
Pur
The results presented thus far strongly implicate Pur
as playing a role in regulating the expression of the
4 subunit gene via interactions with E1. To determine whether Pur
(or an
immunologically related protein) is present in a more physiological
context, we carried out supershift EMSA using a monoclonal antibody
against Pur
and crude nuclear extracts prepared from bovine brains.
The extracts were incubated with one of three radiolabeled E1 probes (the coding strand, the noncoding strand, or double-stranded E1) and an
anti-Pur
monoclonal antibody (24) prior to electrophoresis (see
"Experimental Procedures"). Both single- and double-stranded probes
were used to gain more insight into the interactions between Pur
and
E1; Pur
has been shown to bind to both single- and double-stranded DNA targets (19). As shown in Fig. 6, incubation of the
coding strand probe (ssc) with nuclear extract led
to the appearance of a single DNA-protein complex (lower
arrow). The mobility of this complex was unaltered by the addition
of anti-Pur
antibody or a preimmune serum. Similarly, incubation of
the noncoding strand probe (ssnc) or the
double-stranded probe (ds) with nuclear extract led to the
appearance of one major DNA-protein complex (Fig. 6, upper arrow). These complexes migrated more slowly than that seen with the coding strand probe raising the possibility that different proteins
can interact with the two strands. In support of this hypothesis is the
observation that the addition of anti-Pur
antibody led to a slight
decrease in the mobilities of the complexes formed with the noncoding
and double-stranded probes (Fig. 6, arrowhead). Although the
decrease seen following addition of anti-Pur
antibody was not as
large as might be expected given the size of the antibody, it may be a
consequence of the overall electrostatic charge of the
DNA-Pur
-anti-Pur
antibody complex, such that the binding of
antibody may not greatly change the overall charge of the DNA-protein complex. Given that separation through native acrylamide gels is based
on charge rather than size, the small change in overall charge upon
binding of antibody may not result in a large decrease in mobility.
Nonetheless, these data suggest the presence of Pur
, or an
antigenically related protein, in bovine brain and further suggest that
different proteins are capable of interacting with the coding and
noncoding strands. It is also possible that the proteins which interact
with the coding strand also interact with the noncoding strand and
double-stranded DNA but in conjunction with Pur
, resulting in a
slower migrating DNA-protein complex. Alternatively, different proteins
may interact with the two strands and the affinity of the protein for
the noncoding strand is higher than the affinity of the protein for the
coding strand such that when the double-stranded probe is used, it
forms a complex with the noncoding-strand-binding-protein. Further
analyses should help resolve these questions.
In summary, we used a combination of phosphocellulose, DNA-cellulose,
and DNA sequence-specific affinity chromatographies to purify the E1
binding activity approximately 19,000-fold from bovine brains. There
are four major polypeptides in the most purified fraction. One of the
proteins has been identified as the bovine homologue of the
transcriptional regulatory factor, Pur. The identities of the other
three proteins remain unknown but are the focus of current efforts. In
addition, the functional role Pur
plays in
4 subunit gene
expression is under investigation. In this regard, it is interesting to
note that Pur
has been shown to interact with the retinoblastoma
protein, an important regulatory factor known to be involved in cell
cycle regulation and the expression of specific genes required for
cellular differentiation (26). The molecular weight of the
retinoblastoma protein is approximately 110 kDa, raising the
possibility that NARP114 may be the bovine retinoblastoma protein.
Finally, we previously showed that Sp1 and Sp3 are involved in
4
gene expression via their interactions with E2, which is located
immediately downstream of E1 (14).2 We
hypothesized that Sp1 and Sp3 may interact with the proteins that bind
to E1 (14).2 With the identification of NARP43 as Pur
,
we are now in a position to begin testing that hypothesis.
We thank Drs. Wen-Hwa Lee and Phang-Lang Chen
for the GST-Pur construct, Dr. Edward M. Johnson for anti-Pur
antibodies, Drs. William Lane and John Neveu of the Harvard
Microchemistry Facility for amino acid sequencing, the folks at
Kiolbassa Provision Company for bovine brains, and our colleagues in
the Department of Molecular Medicine (in particular, Drs. Catherine
Bigger, Steve Britt, Eileen Lafer, Irena Melnikova, Eduardo Montalvo
and Kondury Prasad) for many useful discussions and reagents.