| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Cancer (H.L.L., L.W., D.A.Z.), Biophysics (M.H.S.), and Biologics Research (J.M.) Berlex Biosciences Richmond, California 94804
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
(TGF), and 52-kDa cathepsin D (2, 4). However, a seemingly
paradoxical effect of E was also observed in our studies with these
cells; instead of acting as a proliferative stimulus as it does in
ER-positive (ER+) breast carcinomas, E treatment of the B5-ER cells
resulted in marked growth inhibition (5). We therefore proposed that E
could have both positive and negative effects on growth, depending upon
the cellular context of the expressed ER (4). In this regard, recent
studies (6, 7) that investigated the proliferative status of ER+ and
ER-negative cells in the normal human breast by in situ
immunohistochemical staining for the ER, PR, and Ki67 demonstrated that
it is primarily the ER-negative and not the ER+ cells that proliferate.
It is conceivable that exposure of the normal breast to E prevents the
proliferation of the ER+ cells but induces them to produce paracrine
growth factors that enhance the growth of nearby ER-negative cells.
Indeed, the response of the B5-ER cells to E treatment is growth
inhibition (5) and increased expression of growth-modulatory factors
such as TGF and 52-kDa cathepsin D (2). In contrast to the normal
mammary gland, in ER+ breast tumors, ER positivity and proliferation
are significantly associated (6). This suggests that escape of ER+
cells in the normal breast from an E-mediated inhibitory effect on
growth could be one step in the development of malignancy. Elucidation
of the mechanisms regulating the antiproliferative activity of E in the
B5-ER cells should aid our understanding of the means by which ER+
breast tumors evade this control.
Our studies using this B5-ER cell model have revealed that, like E, the
partial ER agonist-antagonist, 4-hydroxy tamoxifen (HT), also inhibited
cell growth (5), while the full ER antagonist, ICI164384 (ICI) did not.
These data suggested that the effect of HT on B5-ER cell growth is a
result of its partial estrogenic activity. This partial estrogenicity
has been implicated in the emergence of some tamoxifen-resistant breast
carcinomas (8), as well as in tamoxifen’s beneficial effects in
protecting postmenopausal women from cardiovascular disease and
osteoporotic bone loss (8). As expected for a partial ER agonist, we
found that only some E-responsive genes were modulated by HT treatment,
i.e. HT increased TGF and pS2, but not 52-kDa cathepsin D
or PR mRNA levels (Ref. 4 and data not shown). Such selective gene
expression regulation by HT has been attributed to promoter- and cell
type-specific differences in the requirements for ER-mediated
transcriptional activation (for review, see Ref. 9). To elucidate these
mechanisms, numerous studies have been performed using the promoters of
several well characterized E and HT-responsive genes and various
mutated ER that are, for example, defective in DNA interaction or in
transcriptional activation. One model to explain the tissue-specific
activity of tamoxifen invoked selective activation of the AP1
transcription factor by HT in uterine, but not breast, cancer cells
(10). In this model, different molecular mechanisms were implicated in
E- and HT-dependent regulation; the activation of AP1 by E was
independent of the DNA-binding ability of the ER, while HT activation
required the intact DNA-binding domain (DBD).
In our previous studies, we determined the ER structural domains that were important for E- and HT- mediated antiproliferative responses in 184B5 cells stably expressing mutant ER. Our results suggested that different mechanisms were involved in the growth-inhibitory effects of E and HT. However, the requirements were not the same as for AP1-dependent transcriptional control; HT-dependent effects could be elicited in cells containing mutant ER that did not have DNA-binding ability, while E effects were dramatically reduced in the same cells.
With the aim of better understanding the mechanism for the E- and HT-dependent antiproliferative effects on B5-ER and to identify genes that could be used to study this novel pathway of HT-mediated gene expression modulation, we employed differential display PCR in the search for E-regulated genes. This strategy resulted in the identification and cloning of a cDNA encoding a protein with homology to members of the B box zinc finger protein family. The hormonal regulation and chromosomal mapping of this gene, as well as the subcellular localization of the encoded protein, are described.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
A combination of cDNA library (i.e. human testis) screening
with this 167-bp cDNA probe and 5'-rapid amplification of cDNA ends
(RACE)-PCR cloning using RNA isolated from the B5-ER cells, as
well as human mammary gland and skeletal muscle, led to the isolation
of a 2568-bp cDNA and several overlapping cDNA clones that correspond
to the sequence provided in Fig. 1A
. DNA sequence
analysis (11) revealed a 564-codon open reading frame. The ATG start
codon is in an adequate context for translation initiation (12), and
stop codons are present in all three upstream reading frames. A
putative polyadenylation signal (i.e. AATAAA) is found at bp
2549 of this cDNA clone. These results and the similarity between the
size of the cDNA and the 2.6-kb mRNA expressed in B5-ER cells (Fig. 2) suggest that this cDNA may be
full-length. The predicted protein has a calculated molecular
mass of 64 kDa and an isoelectric point of 5.2. No signal
peptide sequence or nuclear localization signal was identified.
|
|
. Like ATDC, efp, TIF1 (16), Kap1 (17), and PML, EBBP
contains two B boxes (called B1 and B2). Amino acid sequence analysis
using the Coil Scan program (11, 18) predicted the existence of two
-helical regions within EBBP immediately downstream from the B boxes
(i.e.,Ser164 to Ser203 and
Ala244 to Met272).
The B box motif and a predicted coiled-coil domain are found in all of
the B box family proteins, some of which also contain a RING zinc
finger (19) or a B30.2 homology region (20, 21) (Fig. 1C). The EBBP
does not have a RING finger but contains sequences that correspond to
the consensus sequence for the B30.2 domain (20). It is noteworthy that
a subset of B30.2 domain-containing proteins do not have B boxes (Fig. 1C) and that the B30.2 domain of EBBP is most homologous to those in
non-B box family members, i.e. stonustoxins and ß and
BT2 (with sequence identities of 32%, 34%, and 32%, respectively;
data not shown).
EBBP Tissue Distribution
The tissue specificity of EBBP expression was assessed by Northern
blot analysis of poly(A)+ RNA isolated from various human
adult and fetal tissues. As shown in Fig. 2A, two EBBP transcripts of
2.6 kb and 2.2 kb were detected in the adult tissues tested. The 2.6-kb
mRNA was detected in most of the tissues with the highest levels being
found in testis, ovary, small intestine, colon, placenta, heart, and
mammary gland (Fig. 2A). The expression pattern of the 2.2-kb species
is different from that of the 2.6-kb mRNA; the highest levels of this
form are found in skeletal muscle, heart, and testis. No EBBP mRNA was
detectable in liver or peripheral blood lymphocytes. Comparison of
these data with those obtained from analysis of four fetal tissues
(Fig. 2B) revealed that EBBP was more highly expressed in the fetus
than in the corresponding adult tissues. This is most apparent for the
brain, where EBBP expression is barely detectable in the adult. There
is also an 
3 kb EBBP mRNA detected in fetal tissues, most noticeably
in the brain.
Chromosomal Localization
Using the EBBP cDNA as a hybridization probe for Southern blot
analysis of DNA isolated from a panel of human-rodent somatic cell
monochromosomal hybrids, EBBP was localized to human chromosome 17
(Fig. 3A). The EBBP probe also hybridized
to mouse and hamster DNA, suggesting the existence of rodent homologs
of this gene. The chromosomal location of the EBBP gene was mapped by
two-color fluorescence in situ hybridization using a P1
genomic clone containing the EBBP-coding sequences and a chromosome 17
centromere-specific probe (Fig. 3B). By this analysis, the EBBP gene
was determined to be at a position that is 26% of the distance from
the centromere to the telomere of chromosomal arm 17p, in a region that
corresponds to band 17p11.2.
|
)
or in the HepG2 human hepatocellular carcinoma cell line (data not
shown), the HepG2 cell line was selected for the transfection with the
EBBP expression constructs. Northern blot analysis demonstrated the
presence of EBBP mRNA in the EBBP-transfected cells but not in the
control vector-only transfectants (Fig. 4A). Cell extracts prepared from these
cells were analyzed for protein expression by immunoblotting using a
polyclonal anti-EBBP peptide antibody (Fig. 4B). Two protein bands of
approximately 64 and 70 kDa were detected in the EBBP-transfected
cells, but were absent in the vector-only controls. The majority of the
translated protein product was found as the 70-kDa band. The presence
of this higher molecular mass band, in addition to the band of the
predicted mass of 64 kDa, may be due to posttranslational modification.
The incorporation of the eight amino acid FLAG epitope tag resulted in
the same relative expression of the two proteins, but of a slightly
larger size than the untagged proteins, as expected. When the M2
monoclonal anti-FLAG antibody was employed in this analysis, only
the FLAG-tagged EBBP was detected (data not shown).
|
) and the M2 monoclonal anti-FLAG
antibody (Fig. 5D) were used to detect the EBBP protein. Both
antibodies detected the EBBP protein in the cytoplasm of the positive
cells; only weak, diffuse staining was observed using the isotype
control antibodies (Fig. 5, A and C). Similar cytoplasmic localization
was observed in breast and ovarian cancer cells transfected with EBBP
(data not shown). Biochemical fractionation procedures and Western blot
analysis confirmed the cytoplasmic localization of the majority of the
transfected EBBP (data not shown).
|
; B5-ER clones E-4 and E-23), but
had no effect on the vector-only transfected clone (Fig. 6A; B5 clone
V-1) or the parental 184B5 cell line (data not shown). No stimulatory
effect was seen after treatment with the pure ER antagonist ICI (Fig. 6, A and B), but the effect of E was blocked by cotreatment with ICI at
concentrations that are 10- to 100-fold greater than those of E (Fig. 6B). These data demonstrate that the ER is required for the E-dependent
increase in EBBP mRNA levels.
|
3.5-fold) were not
observed until 24 h (Fig. 6C). To determine whether this
E-dependent regulation required ongoing protein synthesis, studies were
performed with the protein synthesis inhibitor, cycloheximide (CHX;
Fig. 6D). The presence of CHX did not prevent the E-dependent increase
in EBBP mRNA levels at any of the time points tested. All of these data
demonstrate that regulation of EBBP expression is an early, ER-mediated
response that does not require new protein synthesis.
EBBP mRNA Levels Are Increased by HT through a Mechanism That Does
Not Require an Intact ER DNA-Binding Domain
Our previous studies indicated that the partial agonist-antagonist
HT had some estrogenic activities in the B5-ER-transfected cells. Both
E and HT caused antiproliferative effects and induced the expression of
the pS2 gene (4, 5). But, unlike E, HT did not induce PR gene
expression (4). We therefore evaluated the effect of HT treatment on
EBBP mRNA expression. As shown in Fig. 7A
(left side), regulation of EBBP mRNA levels by E treatment
was dose-dependent with maximal effects observed at concentrations of
10-7 M or greater at both the 24- and 72-h
timepoints. In contrast, HT treatment caused an increase in EBBP mRNA
levels only at the 72-h timepoint and with a biphasic dose-response
curve that peaked at 10-8 M and declined at
higher concentrations.
|
(Fig. 7A, right side) and pS2 (data
not shown). TGF mRNA levels increased with increasing concentrations
of both E and HT, with maximal responses observed at 10-7
M and 10-8 M, respectively.
We had previously shown that the antiproliferative effects of E and HT
could be distinguished by different activities of these ER ligands on
B5-ER transfectants bearing mutations in the DBD. In DBD
mutant-containing cells, HT was much more effective than E at eliciting
growth-inhibitory effects (4). We therefore asked whether the
DNA-binding ability of the ER was critical for EBBP regulation by E and
HT. 184B5 cells stably transfected with an ER mutant lacking the second
zinc finger of the DNA-binding domain [DBD-
(4)] were employed in
these analyses. Each of the tested DBD- transfectants exhibits
levels of ER comparable to those found in the WT-ER- transfected cells,
and the relative binding affinities of E and HT for the DBD- ER are
similar to those for the WT-ER (4). As shown in Fig. 7B (left
side), E had only a slightly stimulatory effect on EBBP expression
in the DBD- cells (i.e. 1.7-fold relative to the 2.8-fold
effect on the WT-ER) at the highest dose (i.e.
10-6 M). In contrast, HT (10-7
M) increased EBBP mRNA levels by 3-fold in the DBD-
cells. Moreover, the response to increasing concentrations of HT was
not biphasic as in the WT-ER cells, but was similar to the effects of E
in the WT-ER cells (compare with Fig. 7A, left side). The
HT-mediated increase in EBBP mRNA levels in the DBD- transfectants
was inhibited by cotreatment with the pure ER antagonist ICI (data not
shown), demonstrating that HT binding to the mutant ER is required for
EBBP mRNA expression.
For comparison, we measured the effects of E and HT on TGF gene
expression. In contrast to the observations for EBBP, neither E nor HT
was effective in increasing TGF mRNA levels in the DBD-
mutant-expressing cells (Fig. 7B, right side), consistent with a
requirement for DNA binding in the hormonal regulation of TGF
expression. Other genes, such as PR and pS2, were also unresponsive to
E or HT treatment in the DBD- mutant-expressing cells (4). These
data suggest that EBBP differs from other E and HT-responsive genes in
that an intact ER DBD is important for E, but not for HT-mediated
regulation.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and ß
(34)]. The B30.2 domain of butyrophilin has been suggested to be
involved in milk fat secretion in the breast (35), and three missense
mutations clustering in the B30.2 domain of pyrin have been proposed to
be responsible for familial Mediterranean fever (28). In EBBP, the
B30.2 domain may play an important role in its biological function. The EBBP mRNA is expressed at varying levels in most human tissues by Northern blot analysis, with the exception of liver and peripheral blood leukocytes. Therefore, EBBP tissue distribution is not limited to the classical E target tissues (e.g. ovary, uterus, mammary gland), and its expression is not totally dependent upon E. This observation contrasts with the more restricted expression of efp, the other E-responsive member of the B box protein family, which is very highly expressed in placenta, uterus, and ovary with lower levels in the mammary gland and liver (36).
The EBBP is localized in the cytoplasm of EBBP-transfected HepG2 liver carcinoma cells as well as breast and ovarian cell lines transfected with the same EBBP constructs (data not shown). Although most members of the B box zinc finger protein family are nuclear proteins, they can exhibit cytoplasmic localization in some instances, such as during particular stages in embryonic development [i.e. XNF7 (37)]. Recently, it was shown that the subcellular distribution of RFP is cell type specific and that the RING finger motif was required for its nuclear localization in HeLa cells, while the B box was important for its cytoplasmic retention in NIH3T3 cells (38). Interestingly, the ATDC protein, which has a B box but no RING finger, is also a cytoplasmic protein (39). The absence of a RING finger, signal peptide, and a nuclear localization signal in the EBBP is consistent with its cytoplasmic localization. However, in view of the reported cell type-specific or developmental stage-dependent subcellular localization of the other B box family members, it is possible that EBBP will have a different localization in other cell lines or under altered conditions.
Chromosomal mapping by fluorescence in situ hybridization (FISH) analysis revealed that EBBP is located at chromosome 17p11.2, a region that has no known association with cancer. This locus is, however, associated with several syndromes having clinical manifestations of neurological disorder: Sjorgren-Larsson syndrome (40), Charcot-Marie-Tooth neuropathy-1A (41), hereditary neuropathy with pressure palsies (42), Dejerine-Sottas syndrome (43), and Smith Magenis syndrome (44).
The levels of EBBP mRNA are increased after treatment of B5-ER
cells with either E or the partial agonist-antagonist, HT. The B5-ER
cells employed in our studies are HMEC, which stably express the HEO
cDNA, which encodes an ER protein that has a Gly
Val amino acid
replacement at position 400. This protein is less stable and has a
lower binding affinity for E at physiological temperature than the
wild-type (WT) ER, whose corresponding cDNA is called HEGO (45).
Nevertheless, many of the tested transcriptional activating properties
of the HEO-encoded ER are identical to the HEGO ER. One exception was
reported by Jiang et al (46) concerning tamoxifen effects on
the proliferation of HEO vs. HEGO-transfected MDA231 breast
cancer cells; they found that the HEO form of the ER was more sensitive
to tamoxifen than was the HEGO ER. However, we have found little
difference in the antiproliferative or gene expression-modulatory
effects of E and HT on HEGO vs. HEO-expressing HMEC (data
not shown). The regulation of EBBP mRNA levels by E and HT in the
B5-HEGO transfectant (data not shown) is very similar to that observed
in the B5-ER (i.e. HEO) clones presented here. Due to the
extremely low efficiency of generating stable HEGO- expressing
transfectants (4, 47), the HEO parental background was used to evaluate
the activities of WT or mutant ER described in this study.
We have shown that E-responsive EBBP expression requires the ER since
it does not occur in the parental or vector-only transfected 184B5
cells and is blocked by the pure ER antagonist ICI. Regulation of EBBP
mRNA expression by E is a direct ER-mediated effect that is detected as
early as 2 h post-treatment and occurs in the presence of the protein
synthesis inhibitor, cycloheximide. Our data for HT-dependent EBBP gene
regulation demonstrate kinetics that are delayed relative to those of E
and also suggest that the molecular mechanisms that are involved in
EBBP regulation by E and HT are distinct. Studies of EBBP expression in
B5-ER transfectants containing an ER mutant lacking the second zinc
finger of the DBD (i.e. DBD- cells) revealed that optimal
E-dependent regulation of EBBP requires the ER DBD. In contrast, HT was
able to increase EBBP expression in the DBD- cells to the same
extent as in the WT ER transfectants, suggesting that the intact DBD is
not required for this HT-mediated effect. It is theoretically possible
that the HT responsiveness observed in the DBD- cells is dependent
upon the particular mutation that deletes the entire second zinc
finger. However, our preliminary data reveal a similar difference in
the responsiveness of EBBP gene expression to E and HT treatment in
B5-ER transfected with a different DBD mutant [DBD-X (4)], in which
the first zinc finger is disrupted by mutations in critical cysteine
residues.
Regulation of the transcription of ERE-driven promoters has been
shown to require the DBD of the ER (48, 49). The DBD- mutant is
inactive in binding to consensus ERE (50) and cannot activate
transcription of the ERE-driven reporter construct, pVit-tkcat (50, 51). One interpretation of the markedly reduced activity of E in
modulating EBBP expression in the cells containing the DBD- ER is
that DNA binding is critical for E-dependent regulation. Yet, it is
also possible that the mechanism requires an intact ER DBD but not DNA
binding, as has been described for regulation by E of the interleukin 6
(52, 53) and insulin-like growth factor 1 (54) genes.
In contrast to the observations with E, HT enhanced EBBP expression in
the DBD- transfectants. However, HT did not stimulate expression of
the TGF or pS2 genes in these cells, suggesting that the mechanism
by which HT regulates TGF and pS2 is different from the one used for
EBBP. It is known that E-dependent regulation of some genes does not
require the ER DBD [e.g. GATA-1 (55), retinoic acid
receptor- (56), ovalbumin proximal AP-1 site (57)], but the effect
of HT on such genes has not been reported. Of particular relevance to
our observation are results indicating that E- and HT-dependent
activation of the AP1 transcription factor occurs through different
molecular mechanisms that are distinguished by their requirement for
the DBD (10). In this case, activation by tamoxifen, but not E, was
shown to require an intact ER DBD. The EBBP gene may be the first
example of a gene whose HT-mediated regulation does not depend upon an
intact ER DBD, while E-dependent expression does.
In addition to the difference in the ER domain requirement for EBBP
regulatory activity, E and HT also differed in their dose-response
curves in the WT ER transfectants. E treatment enhanced EBBP mRNA
levels in a dose-dependent manner that plateaued at high
concentrations, while similar doses of HT caused a biphasic response,
i.e. increased EBBP gene expression was observed at low
doses but not at higher concentrations. However, this biphasic response
was not observed in the DBD- cells; the dose-dependent response to
HT in these cells was similar to that of E treatment in the WT ER
transfectants. A recent report has indicated that some transcriptional
coregulatory factors can mediate repression [i.e. N-CoR,
SMRT (58)] of HT-, but not E-regulated gene expression. In view of
this, one possible interpretation of our data is that the second zinc
finger that is deleted in the DBD- construct is important for either
directly binding or maintaining the proper ER structure for interaction
with specific corepressors. We propose that in the WT ER cells, high
concentrations of HT drive formation of more HT-bound ER than at low
concentrations. These HT-bound ER can therefore recruit more
corepressors than at low HT concentrations. This would shift the
balance of repressor-free toward repressor-bound ER and thereby cancel
out any stimulatory effect of HT on EBBP expression. We speculate that
these or similar factors would not interact with the HT-bound DBD-
ER and as a result, EBBP transcription would be stimulated. Such a
regulatory mechanism would also predict that under some physiological
or pathophysiological conditions (e.g. corepressor
scarcity), HT would be a potent enhancer of EBBP expression via the WT
ER even at high ligand concentrations. In this regard, it should be
noted that a biphasic dose-response curve is also characteristic of the
effects of HT on the growth of E-dependent breast cancer cells; HT
stimulates proliferation at low, but not high, concentrations (59, 60).
Progression of ER+ breast cancer to a tamoxifen-resistant phenotype
has been hypothesized to involve cellular loss of corepressors (61).
EBBP may be one example of a gene that is coordinately regulated with
this process. The identification of the actual target(s) of the ER
important for EBBP regulation by E and HT and testing of this proposed
model awaits EBBP promoter isolation and characterization.
One of the aims of our differential gene expression cloning strategy
was to identify potential effectors of the E- and HT-induced
antiproliferative effects on ER-transfected HMEC. The changes in EBBP
gene expression elicited by E in the WT ER and DBD- ER cells are
consistent with the possible involvement of EBBP in growth inhibition.
E inhibits the growth of the WT ER, but not the DBD- ER
transfectants; E treatment enhances EBBP mRNA levels in the WT ER, but
not in the DBD- ER cells. Interpretation of the results obtained
after HT treatment is more complicated. HT significantly inhibits the
growth of both the WT ER and the DBD- transfectants and also
increases EBBP expression in both of these cell lines. These data
appear to suggest a correlation between HT’s effects on EBBP
expression and growth. However, although HT is an effective modulator
of EBBP expression in the WT ER cells, its dose-response curve is
biphasic. Furthermore, a biphasic dose-response relationship was not
observed for the antiproliferative activity of HT (5). Therefore, these
data indicate that there is no direct correspondence between
HT-mediated changes in EBBP mRNA levels and growth inhibition. However,
at this time we cannot exclude the possibility that EBBP plays a role
in the growth-inhibitory process at lower HT doses but not at the
higher concentrations.
In summary, we have identified the EBBP gene as an E and HT-modulated gene in immortal, non-tumorigenic HMEC expressing exogenous ER. Further studies will be required to determine whether EBBP plays a role in preventing the proliferation of E-stimulated ER-positive cells in the normal mammary gland. In addition, these studies have also identified a potentially valuable new model system for studying the regulation of gene expression of mixed ER agonist-antagonists, such as HT. Unlike other E- and HT-responsive genes, the modulation of EBBP gene expression by HT occurs in cells expressing ER that do not have DNA-binding ability.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Cell Lines and Culture Conditions
The human liver hepatoma cell line, HepG2 (ATCC, Rockville, MD),
was routinely cultured in Eagle’s MEM containing 10% FBS (Intergen,
Purchase, NY), and 2 mM glutamine, 0.1 mM MEM
non-essential amino acids, 1 mM sodium pyruvate, and 50
µg/ml gentamycin sulfate. The production and characterization of the
stable WT and mutant ER transfectants of the human mammary epithelial
cell line, 184B5, have been previously described (4). The WT ER
transfectants expressing the HEO cDNA (62) that were employed in this
study are clones E-1, E-4, and E-23; the DBD- ER transfectants
expressing the HE34 cDNA (50) are clones D-1, D-4, and D-7. The 184B5
cell line and its transfected derivatives were routinely cultured in
phenol red-free DFCI-1 medium (63) containing 1% dextran-coated
charcoal (DCC)-treated FBS (prepared as described in Ref. 2 except that
there was no heat-inactivation at 56 C). The medium used for the
transfectants contained 100 µg/ml G418. For studies of gene
expression, cells (5 x 105/100-mm dish) were plated
in phenol red-free -MEM supplemented with 5% DCC-treated FBS, 1.0
µg/ml insulin, 12.5 ng/ml epidermal growth factor, 2.8 uM
hydrocortisone (Sigma), 2 mM glutamine, 0.1 mM
MEM nonessential amino acids, 1 mM sodium pyruvate, and 50
µg/ml gentamycin sulfate. After cell attachment overnight, the
culture medium was exchanged with medium containing the ER ligands
(i.e. E2, HT, ICI) indicated for each
experiment. Initiating treatment with 10 µM cycloheximide
1 h before addition of the test compounds tested the effects of
protein synthesis inhibition. For the DD-PCR experiments, the B5-ER
clone E-1 was treated with either vehicle (0.02% ethanol) or 100
nM E2 for 24 h before RNA isolation.
Unless noted otherwise, Life Technology, Inc. (Gaithersburg, MD)
supplied all cell culture reagents.
Differential Display PCR
Differential display PCR was performed following the methods
described by Liang and Pardee (64). Total RNA was isolated using the
guanidinium isothiocyanate/cesium chloride method (65) and treated with
ribonuclease-free deoxyribonuclease I (Boehringer Mannheim,
Indianapolis, IN). One microgram of DNA-free RNA was used for the first
strand cDNA synthesis carried out at 37 C for 1 h in 50
mM Tris-HCl (pH 8.3), 75 mM KCl, 3
mM MgCl2, 20 µM deoxynucleoside
triphosphates (dNTPs), 1 uM dT12-MN primer (M = G, A,
or C; n = G, A, T, or C) with 5 U of Superscript reverse
transcriptase (Life Technology, Inc.). Two microliters of the first
strand cDNA template were used for PCR in 10 mM Tris-HCl
(pH 8.4), 50 mM KCl, 1.5 mM MgCl, 2.5
µM dNTPs, 10 µCi of [-35S]dATP
(Amersham, Arlington Heights, IL), 1 µM
dT12-MN primer (same primer used in cDNA synthesis), 0.2
M 5'-random primer (GenHunter, Nashville, TN), and 1 U of
Taq polymerase (Perkin-Elmer, Foster City, CA). PCR was
performed in a Perkin-Elmer 4800 cycler as follows: 30 sec at 94 C, 2
min at 40 C, and 30 sec at 68 C for 40 cycles, followed by a 5-min
extension at 72 C. PCR products were analyzed on a 6% sequencing gel,
and differentially displayed bands were eluted from the gel and
reamplified by PCR. The PCR-generated products were isolated and used
in preparation of radiolabeled probes for Northern analysis and cDNA
library screening.
Northern Blot Analysis
The Northern blot preparation and hybridization procedure was
carried out essentially as described (66) except that the RNA was
transferred to Zeta-Probe GT nylon membranes (Bio-Rad, Hercules, CA)
and fixed by UV cross-linking. The samples of poly A+ RNA from human
uterus and mammary gland, as well as multiple tissue Northern blots
(adult and fetal tissues), were purchased from CLONTECH (Palo Alto,
CA). The DNA restriction fragment used to probe Northern blots for
expression of the TGF gene has been previously described (66), and
the EBBP-specific probe is a cDNA fragment encompassing the sequences
between bp 534 and 1390. The blots were either exposed to x-ray films
and quantitated by a computing densitometer (Molecular Dynamics,
Sunnyvale, CA) or to an image plate and analyzed by a phosphoimage
analyzer (Fujix, Fuji Medical Systems, Stanford, CT). Normalization
based upon levels of GAPDH mRNA was also carried out when necessary for
quantitation.
cDNA Library Screening
Approximately 1,500,000 recombinant phage from an oligo
dT-primed human testis cDNA library (in ZAP Express, Stratagene, La
Jolla, CA) were screened with the 167-bp DD-PCR-derived cDNA probe.
Replica filters were prepared following the protocol provided by
Stratagene and hybridized with the cDNA probe as described for Northern
blot analysis. After primary and secondary screening, positive clones
were isolated, converted to phagemids by in vivo excision as
recommended by Stratagene, analyzed by restriction mapping, and
sequenced.
RACE Cloning
5'-RACE was performed using the Marathon cDNA amplification kit
(CLONTECH). One microgram of total RNA isolated from E-treated B5-ER
clone E-1 was reverse transcribed to make first strand cDNA. The first
strand cDNA synthesis, second strand cDNA synthesis, and adaptor
ligation were carried out according to the manufacturer’s protocol. In
the subsequent PCR, two EBBP gene-specific primers
(AG-GSP1:GCGGGCTGCATCCAGGGAGACT, reverse complement of bp 713–734;
AG-GSP2:GAAATACGCAAGTTAGCAATT, reverse complement of bp 2553–2568)
were designed to amplify the 5'-end of the EBBP cDNA and the
full-length EBBP cDNA, respectively, in combination with the Marathon
adaptor primer, AP1. For long-distance PCR, the Expand Long Template
PCR system (Boehringer Manheim) was used as recommended in the
protocol. PCR was performed on the adaptor-ligated B5-ER E-1 cDNA, as
well as on Marathon-ready cDNA from human mammary gland and skeletal
muscle (CLONTECH). The PCR was carried out for 5 min at 94 C, 40 cycles
of 30 sec at 94 C, 30 sec at 55 C, and 4 min at 68 C, followed by a
5-min extension at 68 C. After analysis by agarose gel electrophoresis
and Southern blot hybridization, PCR-amplified products were cloned
into the PCRII/PCR2.1 vector (Invitrogen) and sequenced.
Construction of EBBP Eukaryotic Expression Plasmids
The full-length coding sequence of EBBP was isolated from an
EBBP cDNA clone by PCR. The PCR primers (5'-primer:
TGTATACAAACGCGTACCATGGCTGAGTTGGATCTA-ATGG;
3'-primer: GCGCTAGCTACTAAGGAGCAGTCCCCACCAAGG) were designed to create a
3-bp Kozak’s consensus sequence (ACC) immediately upstream of the ATG
start codon of the EBBP coding sequence, and unique restriction enzyme
cleavage sites at both ends (i.e. AccI and
MluI at 5'-end, and NheI at 3'-end). The PCR
conditions were 5 min at 94 C, 35 cycles of 1 min at 94 C, and 2 min at
72 C, followed by a 5-min extension at 72 C. After subsequent
subcloning into the TA cloning vector PCRII, the sequence of the EBBP
coding region was verified. The EcoRI restriction fragment
containing the full-length coding sequence of EBBP was subcloned into
the mammalian expression vector pSV2neo/CMV(2) to generate pCMV-EBBP
for EBBP protein expression. A synthetic linker-adaptor containing the
FLAG-sequence (67) was inserted into the AccI and
MluI sites upstream of the EBBP coding region in pCMV-EBBP
to create plasmid pCMV-FlagEBBP for expression of a N-terminal
FLAG-tagged EBBP protein.
HepG2 Cell Transfection
Calcium phosphate-mediated transfection of HepG2 cells (1.5
x 106/60 mm plate) was performed as described, using 5.7
µg pCMV-EBBP or pCMV-FlagEBBP DNA/plate. After a 6-h incubation with
the DNA precipitates, the cells were subjected to hyperosmotic shock
for 4 min in 15% glycerol-PBS, rinsed twice with PBS, and refed with
fresh medium. Selection was initiated approximately 48 h
posttransfection by addition of medium containing 100 µg/ml G418;
after 2 weeks, surviving colonies were pooled for further analyses.
EBBP Antibody Generation
A synthetic peptide (TNTTPWEHPYPDLPS) representing the EBBP
protein sequence from amino acid residues 396–410 was conjugated to
keyhole limpet hemocyanin. Polyclonal antibody was generated by
immunizing rabbits with the conjugated peptide after a standard
immunization protocol (68). The anti-EBBP antibody was
affinity-purified from rabbit antiserum by Protein A-fast protein
liquid chromatography.
Western Blot Analysis
Lysates from subconfluent cell monolayers of the pooled HepG2
transfectants were prepared, and samples (100 µg protein) were
fractionated on 10% (wt/vol) SDS-polyacrylamide gels (Novex, San
Diego, CA) and transferred to polyvinylidene difluoride blotting
membranes (Novex). Immunodetection was carried out as previously
described (2), except that the primary antibody was a polyclonal
anti-EBBP antibody (P4) used at 1:1000 dilution or a mouse monoclonal
anti-FLAG antibody (M2) used at 10 µg/ml, and the secondary antibody
was a horseradish peroxidase-coupled goat antimouse/rabbit IgG
(Boehringer Manheim) used at 1:125 dilution.
Immunofluorescent Localization Studies
HepG2 cells were plated (1–2 x 105 cells per
well) in two-well glass chamber slides (LabTek, Naperville, IL). After
overnight incubation, the cells were washed twice with PBS, followed by
a 10-min fixation in 4% formaldehyde in PBS. Cells were rinsed with
PBS and permeabilized by submerging the slides in a -20 C methanol
bath for 4 min, followed by 4 min in a -20 C methanol/acetone (1:1)
bath. To rehydrate, cells were washed in PBS and incubated for 1 h
at room temperature with the indicated primary antibody (rabbit
polyclonal anti-EBBP peptide antibody (1:100), mouse monoclonal
M2-anti-FLAG antibody (25 µg/ml; Eastman Kodak, Rochester, NY), or
mouse or rabbit IgG (1:100; Vector Laboratories, Burlingame, CA). The
cells were sequentially washed for 10 min each with PBS, PBS-containing
Avidin (Avidin/Biotin Blocking Kit; Zymed, South San Francisco, CA),
and PBS. The secondary antibody, goat antimouse, or rabbit IgG (Vector
Laboratories) was diluted 1:100 in PBS-containing Biotin (Avidin/Biotin
Blocking Kit by Zymed) and applied for 30 min after a PBS wash,
followed by a fluorescein isothiocyanate-conjugated
streptavidin (1:200; Molecular Probes, Inc., Eugene, OR) for 30 min.
The stained cells were photographed using a Zeiss Axioplan microscope
equipped with a fluorescein-specific filter.
Chromosomal Localization Studies
A Southern blot containing DNA isolated from human-rodent
somatic cell monochromosomal hybrids was purchased from Oncor
(Gaithersburg, MD). Hybridization using a random primer-labeled probe
prepared from the EBBP cDNA was performed as described for Nothern blot
analysis. FISH studies were performed by Genome Systems (St. Louis,
MO). Purified DNA from an EBBP P1 clone (identified by hybridization to
the EBBP cDNA) was labeled with digoxigenin dUTP by nick translation. A
biotin-labeled chromosome 17 centromere probe was cohybridized with the
EBBP P1 probe in the presence of sheared human DNA to normal metaphase
chromosomes derived from phytohemagglutinin-stimulated
peripheral blood lymphocytes in a solution containing 50% formamide,
10% dextran sulfate, and 2xsaline-sodium citrate essentially
as described (69). Specific hybridization signals were detected by
incubating the hybridized slides in fluoresceinated antidigoxigenin
antibodies and Texas Red avidin followed by counterstaining with
4,6-diamidino-2-phenylindole. Measurements of specifically
hybridized chromosomes 17 were made using digital imaging microscopy as
described (69).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Present address: Molecular and Nuclear Medicine, Lawrence Berkeley
National Laboratory, One Cyclotron Road, Donner Laboratory, Berkeley,
California 94720. 
Received for publication May 12, 1998. Revision received July 7, 1998. Accepted for publication July 23, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 promoter via a novel sequence. Mol
Endocrinol 10:477–487[Abstract]
