Institut für Physiologische Chemie,
Universität München, D-80336 Munich,
Germany,1 and
Eukaryotic Transcription Laboratory, Marie Curie Research
Institute, Surrey RH8 0TL, United Kingdom2
Received 24 October 1997/Returned for modification 2 December
1997/Accepted 17 February 1998
 |
INTRODUCTION |
The expression of the
PHO5 gene in Saccharomyces cerevisiae, which
codes for a secreted acid phosphatase, is strongly repressed in
phosphate-containing media (17). Two transcription factors, the basic helix-loop-helix protein Pho4 and the homeodomain protein Pho2, are required for transcriptional activation of the
PHO5 promoter upon phosphate starvation (20). The
activity of Pho4 is regulated through phosphorylation by a
cyclin-cyclin-dependent kinase complex encoded by PHO80 and
PHO85, respectively (15). Under repressing
conditions, Pho4 is phosphorylated and located predominantly in the
cytoplasm (19). In addition, Pho80 appears to repress Pho4
activity by direct interaction (13).
Deletion analysis of the PHO5 promoter revealed two
regulatory elements, UASp1 and UASp2 (22), to which Pho4 has
been shown to bind in vivo upon phosphate starvation, but not under
high-phosphate conditions (30). The repressed
PHO5 promoter is packaged in a positioned array of
nucleosomes that is interrupted only by a short hypersensitive region
containing UASp1 (1). Upon induction of the gene, a 600-bp
region of the PHO5 promoter becomes hypersensitive to
nucleases, reflecting a profound alteration in the structure of four
nucleosomes (2). Binding of Pho4 to both UASp1 and UASp2 is
required for this transition to occur, which appears to be a
prerequisite for transcriptional activation (27).
In contrast, the role of Pho2 in PHO5 regulation is much
less clear. Although Pho2 is strictly required for PHO5
promoter activation, no Pho2 target sites relevant for promoter
activation have been located so far. Deletion of the one Pho2 binding
site that was previously mapped in vitro (31) did not
influence PHO5 promoter activity significantly
(22). The finding that the activation of a heterologous
promoter by a 31-bp oligonucleotide containing UASp1 was fully Pho2
dependent led to the suggestion that Pho2 acts as a
trans-acting factor without binding to DNA (24).
Pho2 is a pleiotropic effector which is involved in the regulation of a
diverse array of other genes. Together with Swi5, it binds
cooperatively to a regulatory element in the HO promoter (7) and plays a complex role in its regulation
(18). Also named Bas2, Pho2 is involved in the regulation of
HIS4 (3), TRP4 (6), and
certain ADE genes (8). It was recently reported that at the ADE5,7 promoter, Pho2 (Bas2) binds to a site
located immediately adjacent to the Bas1 site, and indirect evidence
suggests that these two proteins bind DNA cooperatively (21,
32).
In this paper, we have dissected the mechanism of Pho2 action at the
PHO5 promoter and provide an answer to the long-standing question of its requirement in PHO5 activation. The strategy
taken is based on our recent in vitro mapping experiments, in which we
demonstrated multiple Pho2 binding sites in the PHO5
promoter. We had also found that Pho2 can bind cooperatively with Pho4
at each Pho4 binding site in vitro (5), raising the
possibility that cooperativity between Pho2 and Pho4 might play a role
in PHO5 activation in vivo. We now show that Pho2 acts
through multiple DNA binding sites and binds to the PHO5
promoter in a cooperative manner with Pho4. Remarkably, two critical
functions for Pho2 have emerged from our experiments, the first being
to recruit Pho4 to the DNA and the second being to enhance its
activation potential once bound to the DNA. At UASp1, recruitment of
Pho4 seems to be the crucial role of Pho2, while at UASp2, it is mainly the second function of Pho2 which is required.
| |
MATERIALS AND METHODS |
Yeast strains and media.
All S. cerevisiae
strains used in this study are isogenic with strain YS18 (MAT
his3-11 his3-15 leu2-3 leu2-112 canr
ura3
5). YS22 contains a disruption of the PHO4
gene, YS19 of the PHO2 gene, and YS27 of PHO4 and
PHO2. Yeast strains were grown in YNB medium (Difco,
Detroit, Mich.) supplemented with the required amino acids
(high-phosphate conditions) or in phosphate-free synthetic medium
(29).
Plasmids.
The PHO2-HIS and PHO4-HIS
expression plasmids were constructed as described previously (5). The
PHO4int-HIS plasmid was created by subcloning
a PHO4int internal fragment into PHO4-HIS.
YEp-Pho4 and Yep-Pho42 have been described by Svaren et al.
(28), and Pho2-VP16 and PHO4int
have been described by Hirst et al. (12). In that paper,
PHO4int was referred to as PHO4200.
The construction of the PHO5-lacZ reporter plasmid was
described previously (26). In the PHO5-lacZ
reporter containing 2× UASp2, UASp1 was replaced by UASp2 as described
for YS70 (30). The PHO5-UAS CYC1-lacZ
reporters were constructed as described previously (24) by
using a 2µm yeast vector containing a CYC1-lacZ gene
fusion (11). A 31-bp promoter fragment extending from 
381 to 351 was used as the UASp1 element (24), and the UASp2
element was a 24-bp oligonucleotide ranging from position 262 to
position 239, corresponding to the Pho4 footprint at UASp2
(31). PHO5 DNA restriction fragments used for
DNase I footprinting were derived from PHO5-lacZ constructs
containing the wild-type or mutated PHO5 promoter.
PHO5 DNA fragments used in gel shift analyses were generated
by PCR with the wild-type or mutated PHO5 promoter as a
template. Mutations of the Pho4 and Pho2 binding sites at the PHO5 promoter were introduced by PCR by the megaprimer
technique (23). Expression and purification of the Pho4-HIS,
Pho4int-HIS, and Pho2-HIS proteins were carried out as described
previously (5).
Functional assays.
-Galactosidase activity measurements
(26), DNase I footprinting and gel shift assays
(5), nuclease digestion of isolated nuclei, and dimethyl
sulfate (DMS) in vivo footprinting with primer 2 (5'-GCCTATTCAATTAACTC) (30) were performed as
previously described (29).
| |
RESULTS |
Pho4 binding and UAS elements at the PHO5
promoter.
In a linker scanning analysis of the PHO5
upstream region, Rudolph and Hinnen had previously detected two regions
which are essential for activation of the promoter by phosphate
starvation (22). Each of these regions was later found to
contain a Pho4 binding site, and they were termed UASp1 and UASp2
(31). Our recent in vitro binding studies revealed that Pho2
binds cooperatively with Pho4 to several sites flanking both UASp1 and
UASp2. In addition, we detected a new, low-affinity Pho4 binding site
located 60 bp downstream of UASp2, and there was cooperativity between
Pho4 and Pho2 at this site as well (5).
Pho4 is bound to the low-affinity site in vivo under derepressed but
not under repressed conditions (data not shown), as we have previously
demonstrated to be true for binding of Pho4 to UASp1 and UASp2
(30). To obtain quantitative information about the
contribution of each of the three sites, which we have shown to bind
Pho4 in vivo, activity measurements of PHO5 promoter
variants containing mutations within each Pho4 site (Fig.
1) were carried out, and the results are
presented in Table 1. Mutation of either of the Pho4 sites at UASp1 or UASp2 leads to a dramatic decrease in
promoter activity (approximately 10-fold), showing cooperative action
of the two sites in the activation process, in agreement with the
previous deletion analysis of the PHO5 upstream region (10, 22). A mutation of the third Pho4 site, in contrast, has a much smaller effect and brings down the activity to about 70% of
the wild-type level. Also, the 14% residual activity of a weakened
promoter variant driven by only UASp2 is not any more strongly
dependent on the third Pho4 binding site than the wild-type promoter. A
variant lacking a functional UASp1 as well as UASp2, but retaining the
third Pho4 site, was practically inactive (1% residual activity
[Table 1]).
View larger version (19K):
[in this window]
[in a new window]
|
FIG. 1.
Mutation of Pho2 and Pho4 binding sites at the
PHO5 promoter. The locations of the Pho4 and Pho2 binding
sites as determined by in vitro footprinting (5) are
indicated by solid and open bars, respectively, the width of the bars
corresponding to the relative affinities of the sites for the factor.
Mutated regions within the Pho2 binding sites are in boxes (M1 to M5),
and the changed nucleotides are shown above the wild-type sequence.
Mutations within the Pho4 consensus sequence are shown below the
wild-type sequence and are referred to as UASp1-mut, UASp2-mut, and
Pho4 Site3-mut.
|
|
We have previously shown that Pho4 is absolutely required for the
transition of the promoter chromatin structure to an open state as a
prerequisite for transcriptional activation (9). It was
therefore conceivable that the third Pho4 site, being occupied in vivo
upon promoter activation, might contribute to nucleosome disruption and
stabilize the open state. We therefore monitored the structure of
nucleosome 2, which contains the third Pho4 binding site, in a
promoter construct in which this site had been mutated by measuring the
accessibility of a ClaI site to the nuclease (1).
Under induced conditions, accessibility to ClaI was
indistinguishable from that of the wild-type promoter (not shown). We
therefore conclude that binding of Pho4 to this third Pho4 site plays
no significant role in the process of chromatin transition and that this site, unlike the Pho4 sites corresponding to UASp1 and UASp2, is
not essential for the activation of the PHO5 promoter.
DNA interaction of Pho2 at UASp1 is required for activity of this
element.
We next turned to the newly discovered Pho2 binding sites
and analyzed their functional role in vivo. To that end, the Pho2 binding sites were mutated individually or in combination (Fig. 1), and
the effects on cooperative DNA binding of Pho2 and Pho4 as well as on
the activity of the mutated promoter variants were determined.
We first tested the function of the strong Pho2 binding sites mapped at
UASp1 which partially overlap the Pho4 site. Two mutations were
introduced in this site, as shown in Fig. 1 and the schematic of Fig.
2. In the shorter mutation (M1), 7 bases
are exchanged in the Pho2-protected region upstream of the Pho4
footprint, while in the longer one (M2), an additional 5 bases
extending into the Pho4 DNase I footprint are mutated. Binding of Pho2
and Pho4 to the mutated promoter fragments was then tested by DNase I
footprinting in vitro. As shown in Fig. 2, M1 results in the loss of
protection in the upstream half of the Pho2 binding region, while M2
brought about complete loss of Pho2 protection, indicating that the
previously mapped Pho2-protected region from 385 to 358
(5) actually represents at least two adjacent Pho2 sites.
Neither of these mutations affects the DNase I footprints of Pho4 (Fig.
2).
View larger version (68K):
[in this window]
[in a new window]
|
FIG. 2.
Effect of the mutations M1 and M2 on Pho2 and Pho4
binding in vitro. DNase I footprinting was performed as described in
Materials and Methods. The upper strand of an SfuI
(206)-BamHI (542) fragment, derived from the wild-type
or mutated promoter, was labeled at the SfuI site. Pho4 or
Pho2 was added as indicated at the top. The regions protected in the
wild-type promoter are indicated on the side. The locations of the M1
and M2 mutations (Fig. 1) within the Pho2 binding site are shown
schematically underneath.
|
|
To see if partial or complete loss of Pho2 DNA binding would result in
the loss of cooperative DNA binding of Pho2 and Pho4, formation of a
ternary complex at the promoter variants was examined by gel shift
experiments, and the results were compared to those for the wild-type
promoter. As shown in Fig. 3, there was
less binary Pho2-DNA complex and concomitant loss of the ternary
complex with the promoter fragment containing the smaller mutation
(M1). Binding of Pho4 itself was not affected by this mutation. When the promoter fragment was used with the larger mutation (M2), there was
no binary Pho2-DNA complex, and the ternary complex was barely
detectable. However, it should be noted that binding of Pho4 itself to
the M2 fragment was slightly impaired (1.5- to 2-times-higher Pho4
concentrations were required for the same amount of binary complex).
View larger version (42K):
[in this window]
[in a new window]
|
FIG. 3.
Pho2 DNA binding is required for cooperativity between
Pho2 and Pho4 at UASp1. The binding reaction and the gel shift assay
were performed as described in Materials and Methods. A labeled 81-bp
PCR promoter fragment (324 to 405) was used, containing either the
wild-type promoter sequence or the M1 or M2 mutation (see Fig. 1), and
is schematically shown at the bottom. The amounts of protein added to
an assay mixture are indicated in arbitrary units. One unit of Pho4 and
Pho2 corresponds to about 5 and 6 ng of protein, respectively, as
determined by sodium dodecyl sulfate gel electrophoresis. The
higher-mobility protein-DNA complex observed with Pho4 added alone
(marked by an arrow) represents proteolytically degraded Pho4 protein
bound to DNA. The positions of the ternary complexes containing either
full-length or degraded Pho4 protein are indicated by asterisks. A
lower-mobility complex with only Pho2 (lane 5) migrates at
approximately the same position as the ternary complex. However, the
presence of a ternary complex with the proteolyzed Pho4 protein (lower
band with asterisk) makes it possible to unambiguously identify the
ternary complex.
|
|
The activities of the mutated promoter variants were measured with
lacZ fusion constructs. As shown in Fig.
4, the activity of the promoter
containing M1 was reduced to about 35%, and that of the one containing
M2 was reduced to about 15%. The residual activity of the M2 promoter
corresponds to the activity of a promoter with a mutated Pho4 site at
UASp1 (Table 1). This result suggests that cooperative binding of Pho2
and Pho4 at UASp1 is crucial for the activity of this element. However,
since disruption of the PHO2 gene brings down promoter
activity to less than 1% (Fig. 8), it is
clear that Pho2 does more than just facilitate binding of Pho4 at
UASp1.
View larger version (42K):
[in this window]
[in a new window]
|
FIG. 4.
Mutations in the Pho2 binding sites differentially
affect PHO5 promoter activity. The activities of the
wild-type PHO5 promoter fused to the lacZ gene
(26) and of promoter variants containing mutations in the
Pho2 binding sites were measured as described in Materials and Methods.
The activities of the mutated promoter variants are expressed relative
to the activity of the wild-type promoter (920 U). The mutations are
schematically shown at the bottom.
|
|
Pho2 binding sites adjacent to UASp2 are also required for full
activation of the PHO5 promoter.
In vitro binding
studies have revealed the existence of Pho2 binding sites closely
adjacent to the Pho4 site at UASp2. Binding of Pho2 to either site
alone gives rise to cooperative DNA binding of the two proteins
(5). Therefore, it was important to check if mutations in
these Pho2 sites would also result in loss of cooperative DNA binding
and a concomitant decrease of promoter activity.
Cooperative DNA binding of Pho2 and Pho4 to the promoter fragment
containing UASp2 and the mutated Pho2 sites upstream and downstream of
the Pho4 site (see schematic in Fig. 5)
was examined by gel shift experiments. As shown in Fig. 5, no binding
of Pho2 to the fragment carrying the M4 and M5 mutations (M4+M5
variant) was observed, nor was any ternary complex detected (compare
lanes 8 to 10 to lanes 3 to 5), clearly showing that Pho2-DNA binding is a prerequisite for cooperative binding of Pho2 and Pho4. Binding of
Pho4 alone is unaffected by the M4 and M5 mutations.
View larger version (42K):
[in this window]
[in a new window]
|
FIG. 5.
Pho2-DNA binding is required for cooperativity between
Pho2 and Pho4 also at UASp2. The binding reaction and the gel shift
assay were performed as described in Materials and Methods. A labeled
109-bp PCR-generated promoter fragment (316 to 208) containing
UASp2 and either the wild-type (wt) sequence or the combined M4+M5
mutation, shown schematically at the bottom, were used. The amounts of
protein added to the assay mixture are listed in arbitrary units (for
details, see the legend to Fig. 3). For the explanation of the arrow
and the asterisks, see the legend to Fig. 3.
|
|
When the activities of the promoter construct containing mutated Pho2
sites upstream and/or downstream of UASp2 were measured, a significant
drop in activity to about 55% was observed for each of the two
single-site mutations (M4 and M5) and to a level of 45% if the two
mutations were combined (Fig. 4). This residual activity was still
significantly higher than the activity of a promoter with a mutated
Pho4 site at UASp2 (7% [Table 1]). These results therefore
demonstrate that Pho2 binding adjacent to UASp2, although contributing
to promoter strength, does not have the same critical importance as at
UASp1.
We have further constructed a promoter variant combining mutated Pho2
sites at both UASp1 and UASp2 (M1+M4+M5). The activity of this
promoter was drastically lowered to less than 10% of that of the
wild-type promoter (Fig. 4) and was similar to the activity of a
promoter containing mutated Pho4 sites at either UASp1 or UASp2 (Table
1). These results show definitively that binding of Pho2 to its target
sites adjacent to UASp1 and UASp2 is required for activation of the
PHO5 promoter. However, one of the strongest Pho2 binding
sites is located between UASp1 and UASp2, and the mutations analyzed so
far had not addressed the contribution of this site by itself to
promoter activity in vivo. Mutation of 6 bp (M3 in Fig. 1) abolishes
interaction with Pho2 within this region, except for a short
Pho2-protected region that still persists immediately upstream (not
shown). The activity of PHO5 promoter constructs containing
this M3 mutation was reduced by about 30% (Fig. 4), showing that
binding of Pho2 to this site also makes some contribution to the
activation of the PHO5 promoter.
Role of the Pho2 cis elements in the chromatin
transition at the PHO5 promoter.
We have previously
shown that the Pho2 binding sites adjacent to UASp1 and UASp2 are
required in vitro for cooperative binding of Pho2 and Pho4 to DNA
(5), and we have now demonstrated that mutation of these
Pho2 sites reduces the activity of the promoter in vivo significantly,
especially when the sites at UASp1 are mutated. The chromatin
transition at the promoter upon PHO5 activation requires
binding of Pho4 to both UASp1 and UASp2 (27) and requires the presence of the Pho2 protein (9). It was therefore
important to determine what consequences the mutations in the Pho2
binding sites would have on the chromatin transition of the promoter
under activation conditions.
The chromatin structure of the different promoter variants with mutated
Pho2 binding sites was analyzed by measuring the accessibility of
restriction sites within nucleosome 2 under inducing conditions. The
results presented in Fig. 6 show a clear
difference in the chromatin structure between the M1 and the M4+M5
variants. By the restriction assay, nucleosome 2 was disrupted in
M4+M5 to almost the same extent as in the wild-type promoter, while
accessibility in the M1 variant dropped to about 20%. Furthermore,
chromatin of the promoter variant containing the combined M1+M4+M5
mutations was closed and nearly indistinguishable by this assay from
the structure of the repressed wild-type promoter. These results
therefore indicate that binding of Pho4 to UASp1 in vivo must be
strongly reduced by mutations of the adjacent Pho2 site (M1), while
binding to UASp2 in the M4+M5 promoter is still productive, although
impaired to a certain degree, as suggested by the closed chromatin of
the M1+M4+M5 promoter compared to the partially open M1 promoter.
View larger version (28K):
[in this window]
[in a new window]
|
FIG. 6.
Chromatin opening at the PHO5 promoter
depends on the Pho2 cis elements adjacent to UASp1 but not
at UASp2. Strains carrying either the wild-type (wt)
PHO5-lacZ plasmid or plasmids with promoter variants were
grown in media containing Pi (+Pi) or not
containing Pi (Pi) as indicated, and nuclei
were prepared. They were digested for 60 min at 37°C in 200 µl of
buffer with 100 U of ClaI or 200 U of HindIII
or XhoI (the M4 mutation introduces an XhoI site
and a HindIII site, whereas the ClaI site is
destroyed). In order to monitor cleavage by the restriction nuclease at
the sites shown in the schematic at the top, DNA was isolated, cleaved
with RsaI, analyzed in a 1% agarose gel, blotted, and
hybridized with a pBR322 RsaI-BamHI fragment
which hybridizes to the region immediately upstream of the
BamHI site. A 1.46-kb RsaI fragment is generated
if the restriction nuclease had been protected, and a fragment about
half that size is generated if the site had been accessible. Analysis
of the wild-type reporter and the M1 promoter is shown at the top.
Accessibility values for the wild-type reporter and M1 promoter as well
as M4+M5 and M1+M4+M5 are shown in the diagram below. Measurements for
M4+M5 and M1+M4+M5 were derived from XhoI and
HindIII digests which gave values that were within 5%
of each other.
|
|
Activation of the PHO5 promoter by Pho2-VP16 requires
multiple Pho2 cis elements in addition to interaction with
Pho4.
It was previously shown that Pho2-VP16 fusion could activate
transcription of the endogenous acid phosphatase only when it was
coexpressed with a DNA binding Pho4 derivative which was inactive by
itself since it lacked an activation domain (12). This
finding demonstrated that Pho2-Pho4 interactions are required for
targeting Pho2-VP16 to DNA. However, it was not clear from these
results whether the Pho2-VP16 fusion was directly bound to DNA or
whether protein-protein interactions with Pho4 were sufficient to
recruit the hybrid Pho2 molecule to the promoter. To address this
question, we have measured the activity of Pho2-VP16 in the presence of Pho42, a transcriptionally inactive derivative of Pho4
(28) in a pho2 pho4 strain, by using
PHO5 promoter constructs with mutated Pho2 binding sites. In
agreement with previous findings, Pho2-VP16 and Pho42 together, but
not individually, activated the wild-type PHO5 promoter as
shown in Fig. 7A. Activation not only
required the Pho4 protein derivative but also required Pho4 binding
sites. The level of activation was significantly reduced after mutation
of the Pho2 binding sites adjacent either to UASp1 or to UASp2 (Fig.
7B) and closely paralleled the results obtained when activation of the
same mutant promoters by Pho4 was tested (see Fig. 4). The most
dramatic effect was observed with multiple Pho2 binding site mutations.
These results further support the conclusion that direct Pho2-DNA
contacts are required in addition to Pho2-Pho4 interactions to activate
the PHO5 promoter and that the multiple Pho2 binding sites
mapped in vitro are functional in vivo.
View larger version (27K):
[in this window]
[in a new window]
|
FIG. 7.
Pho2 cis elements are required for activation
of the PHO5 promoter by Pho2-VP16 and a transcriptionally
inactive Pho4 derivative. (A) Activation of the wild-type
PHO5 promoter or a promoter variant with a mutated Pho4 site
at UASp2 (UASp2-M) by Pho42 and/or Pho2-VP16 as measured in YS22
(pho2) or YS27 (pho4 pho2). (B) Activation of the
PHO5 promoter variants containing mutated Pho2 binding
sites, shown schematically at the bottom, by Pho42 and Pho2-VP16
expressed together in YS27 (pho4 pho2).
|
|
Overexpression of Pho4 relieves the requirement for the Pho2
cis elements.
We have previously shown that
overexpression of Pho4 in a pho2 strain restores its ability
to disrupt the nucleosomes at the PHO5 promoter, indicative
of productive binding of Pho4 to the UAS elements. However, under such
conditions, acid phosphatase activity in the mutant was much lower than
the level in the wild type (9). This finding raises the
question to what extent overexpression of Pho4 in a PHO2
strain would relieve the requirement of the Pho2 cis
elements. As shown in Fig. 8,
overexpression of Pho4 increases the activity of the promoter with the
mutations in the Pho2 sites at both UASp1 and UASp2 (M1+M4+M5) from 9%
to 65 to 70% of the wild-type level. Furthermore, the activities of
promoter variants containing mutated Pho2 sites only at UASp1 or at
UASp2 reached essentially wild-type levels in the presence of
overexpressed Pho4. On the other hand, overexpression of Pho4 measured
in parallel with a wild-type reporter in a pho2 background
gave less than 10% activity (Fig. 8). Therefore, the requirement for
Pho2 binding sites in cis can be relieved by overexpression
of Pho4 to a much greater extent than the requirement of Pho2 as a
trans-acting factor. This finding suggests a more complex
role of Pho2 in the activation process at the PHO5 promoter
than merely facilitating binding of Pho4 to its target sites.
View larger version (32K):
[in this window]
[in a new window]
|
FIG. 8.
Overexpression of Pho4 relieves the requirement for Pho2
cis-acting elements. Activation of PHO5 promoter
variants containing mutated Pho2 cis elements, indicated
schematically at the bottom, was measured in a wild-type (wt) strain
(YS18) and in the same strain expressing Pho4 from a multicopy plasmid.
The activation of a wild-type reporter in a pho2 strain
(YS19) is shown on the right. 2µ, 2µm plasmid.
|
|
A mutant Pho4 protein lacking the Pho2 interaction domain is
defective in cooperative DNA binding with Pho2.
By the use of the
two-hybrid system, Pho4-Pho2 interactions had been demonstrated in
vivo, and the Pho4 segment essential for this interaction was mapped. A
Pho4 protein lacking amino acids 200 to 247 (Pho4int) failed to
interact with Pho2 (12). To test the importance of specific
Pho4-Pho2 interactions for cooperative DNA binding of the two proteins,
gel shift experiments with Pho4int were performed. As shown in Fig.
9A, no significant cooperativity was
observed with Pho4int and Pho2 when binding to a promoter fragment
containing UASp1 was tested. In contrast, strong cooperativity was
observed with wild-type Pho4 and Pho2 in the same experiment (Fig. 3).
A certain degree of cooperativity between Pho4int and Pho2 was
retained when a promoter fragment containing UASp2 with the adjacent
Pho2 binding sites was analyzed, but cooperativity was significantly
reduced compared to that of wild-type Pho4 (Fig. 9B). Similar results
were obtained with UASp2 promoter fragments containing only the 5' or
the 3' Pho2 binding site (not shown). These data show that specific
protein-protein interactions are important for cooperative binding of
the two proteins to the PHO5 promoter.
View larger version (52K):
[in this window]
[in a new window]
|
FIG. 9.
Cooperative DNA binding with Pho2 is largely abolished
with a Pho4 variant lacking the Pho2 interaction domain. The binding
reaction and the gel shift assay were performed as described in
Materials and Methods. Proteins were added individually or in
combination in the amounts indicated (arbitrary units as in Fig. 3,
with 1 U of Pho4int corresponding to about 5 ng of protein) to a
labeled 81-bp PCR-generated promoter fragment (324 to 405)
containing UASp1 and the overlapping Pho2 sites (A), or to a labeled
109-bp PCR-generated fragment (316 to 208) containing UASp2 and the
adjacent Pho2 sites (B) (see schematics at the bottom). The arrows mark
binary complexes derived from proteolytically degraded Pho4 (solid
arrow) and Pho4int (broken arrow). Ternary complexes with
full-length Pho4 and its degradation product are marked by asterisks
and those for Pho4int are marked analogously by dots.
|
|
The Pho4 derivative lacking the Pho2 interaction domain can
activate UASp2 but not UASp1.
We next wanted to test the ability
of Pho4int to activate the PHO5 UAS elements in order to
see if loss of cooperative DNA binding of Pho4 and Pho2 correlates with
transcriptional activation. Our activity measurements of the
PHO5 promoter in vivo suffer, however, from the fact that
both UAS elements must synergize for activity, and differential effects
on just one of the elements would be hard to detect. We therefore
introduced each UAS element into a lacZ reporter plasmid
upstream of a CYC1 minimal promoter and measured the
activity of these constructs with Pho4int and full-length Pho4. It
turned out that Pho4int was completely unable to activate UASp1,
whereas significant activation of UASp2 was measured, which approached
the level obtained with full-length Pho4 (Table 2). Activation of UASp2
by Pho4int is not due to the residual cooperativity with Pho2
observed in vitro (see above), since the same level of activation was
found in the absence of Pho2 (not shown). These results demonstrate
that there is a significant difference between the two UAS elements:
Pho4int can bind to UASp2 and activate transcription, whereas it
cannot bind to UASp1, demonstrating that interaction with Pho2 at this
UAS element is stringently required for Pho4 binding. In the light of
these findings, the poor activation of the PHO5 promoter
with Pho4int (Table 2) can be explained by its inability to bind
(and activate) UASp1, since, as shown before (Table 1), a UASp1
mutation cripples the promoter.
In order to examine activation of UASp2 by Pho4int and full-length
Pho4 in the context of the PHO5 promoter, we decided to replace UASp1 in the PHO5 promoter and to construct a
variant containing 2× UASp2. As expected, this promoter variant was
activated significantly by Pho4int and again was activated in a
Pho2-independent manner (Table 3).
However, activation of this reporter by full-length Pho4 is still
higher in a PHO2 background. This might due to an effect of
Pho2 on the ability of Pho4 either to bind DNA even better, to
transactivate, or both. To address this question, we used in vivo
footprinting to determine the occupancy of the newly introduced UASp2
by using either full-length Pho4 or Pho4int. Unlike the native UASp2
element, the new one is accessible in the nucleus also under repressing
conditions (28), since it resides in a constitutively
hypersensitive chromatin region of the promoter (1). Binding
of full-length Pho4 is indeed improved in the presence of Pho2 (compare
lanes 2 and 4 in Fig. 10). However, even in the absence of Pho2, there was strong binding of Pho4 as well
as Pho4int. As expected, binding of Pho4int is not increased in
the presence of Pho2 (not shown), consistent with the activation results (Table 3).
View larger version (35K):
[in this window]
[in a new window]
|
FIG. 10.
DMS footprint analysis of Pho4 binding to UASp2.
Binding of wild-type Pho4 and Pho4int to the upstream UASp2 element
in the PHO5 promoter variant in which UASp1 was replaced by
UASp2 (see schematic) was examined in YS22 (pho4) or YS27
(pho2 pho4). For details of DMS footprinting and the primer
used (arrow in the schematic), see Materials and Methods.
|
|
Importantly, the footprints were practically identical for Pho4 and
Pho4int (lanes 2 and 3 in Fig. 10), indicating that both molecules
bound with similar efficiency to this UAS element. Since much higher
activity was measured under these conditions with Pho4int compared
to that with full-length Pho4, this must mean that the ability of Pho4
to transactivate can be positively modified. A possible explanation is
the presence of a repressive domain in Pho4 that is at least partially
removed in Pho4int and that is counteracted by Pho2 under
physiological conditions. A similar role of Pho2 in enhancing the
activation potential of Pho4 was recently proposed by Shao et al.
(25) using a different approach.
In conclusion, we cannot unambiguously discriminate to what extent the
increased activation by wild-type Pho4 through UASp2 in the presence of
Pho2 is due to improved DNA binding as opposed to enhanced
transcriptional activation. However, our results with Pho4int show
that Pho2 also has an effect on the ability of Pho4 to transactivate.
Furthermore, there is a clear difference between the two UAS elements
in this respect which had previously not been recognized. At UASp1,
binding of Pho4 is the limiting step for which Pho2 is stringently
required, while at UASp2, Pho2 plays much less of a role in affecting
binding but instead exerts a novel effect by improving the activation
potential of Pho4.
| |
DISCUSSION |
The strict dependence of PHO5 promoter activation on
Pho2 has been known for over 25 years (20), but clues to the
underlying mechanism have emerged only recently. Our recent in vitro
studies revealed the existence of multiple Pho2 binding sites at the
PHO5 promoter in close vicinity to the Pho4 binding sites.
In addition, highly cooperative DNA binding of Pho2 and Pho4 was
demonstrated at UASp1 and UASp2 (5). In this paper, we have
examined the functional relevance of the newly mapped Pho2 sites and
have asked whether the cooperativity between the two proteins plays a
role in the activity of the PHO5 promoter in vivo. To
address these questions, we have constructed a series of different
promoter variants with single or multiple mutations in the Pho2 sites.
The requirement of UASp1 and UASp2 for chromatin disruption at the
PHO5 promoter and subsequent transcriptional activation had
previously been demonstrated by deletion analyses of the promoter (10, 22) and was confirmed here by analyzing the effects of targeted mutations. In contrast to these two elements, the recently mapped low-affinity Pho4 site (5) does not play a crucial
role in the activation process. Although it is occupied in vivo under derepressed conditions, it is dispensable for chromatin disruption (data not shown), and makes only a small contribution to overall promoter activity (Table 1). We have therefore focused our
investigations on the Pho2 sites adjacent to UASp1 and UASp2.
Pho2 is involved in the activation of the PHO5 promoter
through cooperative DNA binding with Pho4.
Mutation of the Pho2
binding sites adjacent to UASp1 and UASp2 results in the loss of
cooperative DNA binding of Pho2 and Pho4 in vitro (see Fig. 3 and 5).
The finding that mutation of any individual Pho2 binding site caused a
moderate to strong reduction in PHO5 promoter activity
indicates that cooperative binding of the two proteins is also
essential for the activity of the PHO5 promoter in vivo.
Furthermore, a dramatic effect was obtained by the combined mutation of
the Pho2 sites adjacent to both UASp1 and UASp2, which was similar to
the effect of mutating UASp1 or UASp2 itself (Table 1 and Fig. 4). The
functional importance of the Pho2 binding sites in vivo was further
confirmed in experiments in which the PHO5 promoter was
activated by a Pho2-VP16 hybrid in the presence of a transcriptionally
inactive Pho4 derivative that could still bind DNA. Promoter variants
with mutated Pho2 sites could not be efficiently activated by Pho2-VP16
(Fig. 7B). Taken together, these results provide the first direct
evidence that Pho2 is involved in the activation of the PHO5
promoter as a sequence-specific DNA-binding protein. At the same time,
they show that binding of Pho2 to DNA requires interactions with Pho4, because in the absence of the Pho4 derivative or a Pho4 binding site,
Pho2-VP16 is transcriptionally silent.
The properties of Pho2 are typical of the family of homeodomain
proteins to which Pho2 belongs. In many cases, such proteins bind DNA
at multiple sites with relatively low sequence specificity in vitro and
are thought to gain their selectivity through protein-protein interactions with other factors (16). The collected evidence therefore strongly indicates that Pho2 contributes to PHO5
activation as a DNA-binding factor which binds to specific sequences at
the PHO5 promoter cooperatively with Pho4 (5).
UASp1 and UASp2 differ in their dependence on Pho2 cis
elements.
Prevention of Pho2 interactions with its target sites
results in the loss of cooperative DNA binding of the two proteins and leads to the progressive weakening of the PHO5 promoter.
However, interference with Pho2 binding around UASp1 has a stronger
effect than that around UASp2, as borne out in decreased promoter
activity on the one hand and in its consequences on chromatin opening
on the other. UASp1 is stringently required for the disruption of the
four nucleosomes at the promoter (10). Similarly, the M1 mutation, in which part of the Pho2 site next to UASp1 is mutated, strongly impairs chromatin opening, and accessibility of the
ClaI site in nucleosome 2 changes from around 90% to
about 20% (Fig. 6). The M2 mutation that eliminates Pho2 binding
entirely has an even stronger effect on activity and chromatin opening
(not shown). However, this mutation extends into the region at which the Pho2 and the Pho4 sites overlap and leads to a slight decrease in
Pho4 binding in vitro even in the absence of Pho2 and was therefore not
further investigated.
In contrast, mutation of the two Pho2 sites adjacent to UASp2 (M4+M5)
affects promoter activity in vivo less, and nucleosome 2 is still
almost completely disrupted. There is a measurable effect of the M4+M5
mutation in the chromatin assay, however. When it is combined with the
M1 mutation, chromatin is almost fully closed in contrast to the
partial opening of the promoter with the M1 mutation. This indicates
that binding of Pho4 to UASp2 in the M4+M5 promoter is not as strong as
that in the wild-type promoter but is still sufficient to disrupt the
nucleosome (Fig. 6). Therefore, it seems that in contrast to UASp1,
binding of Pho4 to UASp2 is improved by cooperative interactions with
Pho2 but is not absolutely dependent on them.
The persistence of nucleosome 2 in a pho2 strain
(9) precludes directly assaying the occupancy of UASp2 by
Pho4, since we have shown that the nucleosome prevents binding of Pho4
to its target site (30). However, construction of a
PHO5 promoter variant with two UASp2 elements, one replacing
UASp1, has made it possible to directly determine the binding of Pho2
to UASp2 now located in a nucleosome-free region. With this construct, we could demonstrate that Pho4 indeed binds strongly to UASp2, even in
the absence of Pho2 (Fig. 10), in agreement with the chromatin data.
A Pho4 derivative lacking the Pho2 interaction domain clearly
distinguishes between UASp1 and UASp2.
It was previously
demonstrated that a Pho4 derivative lacking amino acids 200 to 247, Pho4int, is unable to interact with Pho2 in the two-hybrid system
(12), and Pho4int is indeed defective in cooperative DNA
binding with Pho2 in vitro (Fig. 9). Furthermore, Pho4int can
activate PHO5 transcription only very poorly (about 6% of
the level attained with full-length Pho4), supporting the importance of
cooperativity between Pho2 and Pho4 in physiological activation.
However, activity measurements revealed a striking difference between
the two UAS elements when they were tested out of context. UASp2 was
significantly activated by Pho4int, while UASp1 was not activated at
all. Activation of UASp2 with Pho4int was completely Pho2
independent, showing that residual slight cooperative binding of
Pho4int and Pho2 observed in vitro (Fig. 9B) plays no role in vivo.
Consistent with this difference between UASp1 and UASp2, the
PHO5 promoter variant with UASp1 replaced by UASp2 shows
significant activation with Pho4int in the absence of Pho2.
Surprisingly, under these conditions (i.e., in the absence of Pho2),
wild-type Pho4 activates this construct very poorly, although in terms
of binding to UASp2 it is indistinguishable from Pho4int. This
result throws a new light on transactivation of Pho4 and makes it
likely that Pho4 is negatively regulated by an internal repressive
domain in its ability to transactivate. The simple deletion of the Pho4
segment interacting with Pho2 is sufficient to at least partially
relieve this repression, something otherwise accomplished by
interaction with Pho2.
The Pho2-independent activation of UASp2 resembles that of the recently
described Pho4 mutants containing deletions in the basic region (amino
acids 252 to 265), which is proposed to mediate functional interactions
with Pho2 (25). These deletions result in Pho2-independent
activation of the GAL1 promoter by a Gal4(DBD [for DNA
binding domain])-Pho4 hybrid protein, while full-length Pho4 fused to
the Gal4 DNA binding domain requires Pho2 for activation. Since a Gal4
DNA binding domain was used, a role of Pho2 in the ability of Pho4 to
bind DNA was not addressed. The authors propose a model in which the
Pho4 activation domain interacts with the basic region and is thereby
masked. The role of Pho2 would then be to disrupt this internal
interaction, expose the activation domain, and generate a
transcriptionally competent molecule. In the framework of this model,
the deletion of amino acids 200 to 247 in our experiments, which
results in the inability of Pho4 to interact with Pho2, would destroy
the internal interaction in the Pho4 protein, thus exposing the
activation domain and in effect bypassing the Pho2 requirement.
A second role for Pho2 can also explain our finding that the loss of
promoter activity due to mutation of Pho2 cis elements can
be almost fully compensated for by overexpression of Pho4 in a
PHO2 strain, while the same is not true for the loss of
promoter activity due to elimination of Pho2 itself. These results show that the presence of Pho2 in trans results in higher
transcriptional activity than that measured in a pho2
strain, providing further support for the additional role of Pho2 in
transcriptional activation.
Pho2: a pleiotropic factor in yeast.
Pho2 is involved in the
regulation of several genes, and the common principle in all cases so
far seems to be that it interacts with gene-specific factors. At the
HO promoter, a Pho2 binding site is located next to a Swi5
binding site, and it was shown that the two proteins bind to their
sites cooperatively in vitro (7). However, recent in vivo
data indicate that the role of Pho2 in HO regulation is
complex (18). In the case of the HIS4 promoter, a
Pho2 (Bas2)-protected region largely overlaps the Bas1 footprint.
Although Bas2 and Bas1 can bind to this region simultaneously, no
cooperative interactions between the two proteins were detected
(3). In contrast, at the TRP4 promoter, a Pho2 binding site completely overlaps one of the two Gcn4 binding sites, and
the two proteins were found to bind DNA in a mutually exclusive manner
(6). The role of Pho2 (Bas2) in the activation of the ADE genes was recently investigated (21, 32).
Transcriptional activation of these genes requires the concerted action
of Bas1 and Bas2 and is down-regulated by adenine. From their studies, the authors conclude that Pho2 (Bas2) stimulates both DNA binding and
activation by Bas1 at the ADE5,7 promoter. Interestingly, when a mutant Pho2 protein lacking the DNA binding domain was tested
together with Bas1, it was still partially functional in ADE5,7 activation. This is in contrast to the mechanism of
action of Pho2 at the PHO5 promoter, since our results show
that DNA binding is critically required for Pho2 function. This
difference is also borne out in a recent study by Justice et al.
(14), who showed that mutations in the Pho2 DNA binding
domain almost entirely abolish activation by PHO5 UASp1 and
the UAS elements of the HIS4 as well as the HO
promoter, while an ADE1-lacZ reporter retained 30 to 40%
activity.
Dual role of Pho2 in the activation of the PHO5
promoter.
It is obvious that Pho2 is an exceptional protein when
it comes to the diverse effects it has on cellular metabolism and the means employed to functionally complement the dedicated factor at each
promoter. Cooperative binding with a specific factor appears to be the
primary mechanism, but with the same partner, the requirements for
cooperative binding can be different for different promoters (e.g.,
PHO5 versus PHO8, where Pho4 binding is largely
Pho2 independent [unpublished observations; see also reference
4]), or even more remarkably, at different sites of
the same promoter, as shown here for PHO5 UASp1 and UASp2.
Clearly, the additional role of Pho2 in exposing the activation domain
of the primary activator can be effective only in a case in which
binding of the activator protein is at least to some extent Pho2
independent, as appears to be the case for UASp2 at PHO5.
We thank J. Svaren and Philip Gregory for discussions and
comments on the manuscript and D. Blaschke, A. Schmid, and M. Zavari for expert assistance. We are grateful to D. Stillman for the gift of
PHO2-HIS.
This work was supported by the Deutsche Forschungsgemeinschaft (SFB
190) and Fonds der Chemischen Industrie.
| 1.
|
Almer, A., and W. Hörz.
1986.
Nuclease hypersensitive regions with adjacent positioned nucleosomes mark the gene boundaries of the PHO5/PHO3 locus in yeast.
EMBO J.
5:2681-2687[Medline].
|
| 2.
|
Almer, A.,
H. Rudolph,
A. Hinnen, and W. Hörz.
1986.
Removal of positioned nucleosomes from the yeast PHO5 promoter upon PHO5 induction releases additional upstream activating DNA elements.
EMBO J.
5:2689-2696[Medline].
|
| 3.
|
Arndt, K. T.,
C. Styles, and G. R. Fink.
1987.
Multiple global regulators control HIS4 transcription in yeast.
Science
237:874-880[Abstract/Free Full Text].
|
| 4.
|
Barbaric, S.,
K. D. Fascher, and W. Hörz.
1992.
Activation of the weakly regulated PHO8 promoter in S. cerevisiae: chromatin transition and binding sites for the positive regulator protein PHO4.
Nucleic Acids Res.
20:1031-1038[Abstract/Free Full Text].
|
| 5.
|
Barbaric, S.,
M. Münsterkötter,
J. Svaren, and W. Hörz.
1996.
The homeodomain protein Pho2 and the basic-helix-loop-helix protein Pho4 bind DNA cooperatively at the yeast PHO5 promoter.
Nucleic Acids Res.
24:4479-4486[Abstract/Free Full Text].
|
| 6.
|
Braus, G.,
H. U. Mösch,
K. Vogel,
A. Hinnen, and R. Hütter.
1989.
Interpathway regulation of the TRP4 gene of yeast.
EMBO J.
8:939-945[Medline].
|
| 7.
|
Brazas, R. M.,
L. T. Bhoite,
M. D. Murphy,
Y. X. Yu,
Y. Y. Chen,
D. W. Neklason, and D. J. Stillman.
1995.
Determining the requirements for cooperative DNA binding by Swi5p and Pho2p (Grf10p/Bas2p) at the HO promoter.
J. Biol. Chem.
270:29151-29161[Abstract/Free Full Text].
|
| 8.
|
Daignan-Fornier, B., and G. R. Fink.
1992.
Coregulation of purine and histidine biosynthesis by the transcriptional activators BAS1 and BAS2.
Proc. Natl. Acad. Sci. USA
89:6746-6750[Abstract/Free Full Text].
|
| 9.
|
Fascher, K. D.,
J. Schmitz, and W. Hörz.
1990.
Role of trans-activating proteins in the generation of active chromatin at the PHO5 promoter in S. cerevisiae.
EMBO J.
9:2523-2528[Medline].
|
| 10.
|
Fascher, K. D.,
J. Schmitz, and W. Hörz.
1993.
Structural and functional requirements for the chromatin transition at the PHO5 promoter in Saccharomyces cerevisiae upon PHO5 activation.
J. Mol. Biol.
231:658-667[Medline].
|
| 11.
|
Guarente, L., and M. Ptashne.
1981.
Fusion of Escherichia coli lacZ to the cytochrome c gene of Saccharomyces cerevisiae.
Proc. Natl. Acad. Sci. USA
78:2199-2203[Abstract/Free Full Text].
|
| 12.
|
Hirst, K.,
F. Fisher,
P. C. McAndrew, and C. R. Goding.
1994.
The transcription factor, the cdk, its cyclin and their regulator: directing the transcriptional response to a nutritional signal.
EMBO J.
13:5410-5420[Medline].
|
| 13.
|
Jayaraman, P. S.,
K. Hirst, and C. R. Goding.
1994.
The activation domain of a basic helix-loop-helix protein is masked by repressor interaction with domains distinct from that required for transcription regulation.
EMBO J.
13:2192-2199[Medline].
|
| 14.
|
Justice, M. C.,
B. P. Hogan, and A. K. Vershon.
1997.
Homeodomain-DNA interactions of the Pho2 protein are promoter-dependent.
Nucleic Acids Res.
25:4730-4739[Abstract/Free Full Text].
|
| 15.
|
Kaffman, A.,
I. Herskowitz,
R. Tjian, and E. K. O'Shea.
1994.
Phosphorylation of the transcription factor Pho4 by a cyclin-CDK complex, Pho80-Pho85.
Science
263:1153-1156[Abstract/Free Full Text].
|
| 16.
|
Laughon, A.
1991.
DNA binding specificity of homeodomains.
Biochemistry
30:11357-11367[Medline].
|
| 17.
|
Lenburg, M. E., and E. K. Oshea.
1996.
Signaling phosphate starvation.
Trends Biochem. Sci.
21:383-387[Medline].
|
| 18.
|
McBride, H. J.,
R. M. Brazas,
Y. Yu,
K. Nasmyth, and D. J. Stillman.
1997.
Long-range interactions at the HO promoter.
Mol. Cell. Biol.
17:2669-2678[Abstract].
|
| 19.
|
O'Neill, E. M.,
A. Kaffman,
E. R. Jolly, and E. K. O'Shea.
1996.
Regulation of Pho4 nuclear localization by the Pho80-Pho85 cyclin-CDK complex.
Science
271:209-212[Abstract].
|
| 20.
|
Oshima, Y.
1982.
Regulatory circuits for gene expression: the metabolism of galactose and phosphate, p. 159-180.
In
J. N. Strathern, E. W. Jones, and J. R. Broach (ed.), The molecular biology of the yeast Saccharomyces cerevisiae: metabolism and gene expression. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
|
| 21.
|
Rolfes, R. J.,
F. Zhang, and A. G. Hinnebusch.
1997.
The transcriptional activators Bas1, Bas2, and Abf1 bind positive regulatory sites as the critical elements for adenine regulation of ADE5,7.
J. Biol. Chem.
272:13343-13354[Abstract/Free Full Text].
|
| 22.
|
Rudolph, H., and A. Hinnen.
1987.
The yeast PHO5 promoter: phosphate-control elements and sequences mediating mRNA start-site selection.
Proc. Natl. Acad. Sci. USA
84:1340-1344[Abstract/Free Full Text].
|
| 23.
|
Sarkar, G., and S. S. Sommer.
1990.
The "megaprimer" method of site-directed mutagenesis.
BioTechniques
8:404-407.
[Medline] |
| 24.
|
Sengstag, C., and A. Hinnen.
1988.
A 28-bp segment of the Saccharomyces cerevisiae PHO5 upstream activator sequence confers phosphate control to the CYC1-lacZ gene fusion.
Gene
67:223-228[Medline].
|
| 25.
|
Shao, D.,
C. L. Creasy, and L. W. Bergman.
1996.
Interaction of Saccharomyces cerevisiae Pho2 with Pho4 increases the accessibility of the activation domain of Pho4.
Mol. Gen. Genet.
251:358-364[Medline].
|
| 26.
|
Straka, C., and W. Hörz.
1991.
A functional role for nucleosomes in the repression of a yeast promoter.
EMBO J.
10:361-368[Medline].
|
| 27.
|
Svaren, J., and W. Hörz.
1997.
Transcription factors vs nucleosomes: regulation of the PHO5 promoter in yeast.
Trends Biochem. Sci.
22:93-97[Medline].
|
| 28.
|
Svaren, J.,
J. Schmitz, and W. Hörz.
1994.
The transactivation domain of Pho4 is required for nucleosome disruption at the PHO5 promoter.
EMBO J.
13:4856-4862[Medline].
|
| 29.
|
Svaren, J.,
U. Venter, and W. Hörz.
1995.
In vivo analysis of nucleosome structure and transcription factor binding in Saccharomyces cerevisiae.
Microb. Gene Techniques
6:153-167.
|
| 30.
|
Venter, U.,
J. Svaren,
J. Schmitz,
A. Schmid, and W. Hörz.
1994.
A nucleosome precludes binding of the transcription factor Pho4 in vivo to a critical target site in the PHO5 promoter.
EMBO J.
13:4848-4855[Medline].
|
| 31.
|
Vogel, K.,
W. Hörz, and A. Hinnen.
1989.
The two positively acting regulatory proteins PHO2 and PHO4 physically interact with PHO5 upstream activation regions.
Mol. Cell. Biol.
9:2050-2057[Abstract/Free Full Text].
|
| 32.
|
Zhang, F.,
M. Kirouac,
N. Zhu,
A. G. Hinnebusch, and R. J. Rolfes.
1997.
Evidence that complex formation by Bas1p and Bas2p (Pho2p) unmasks the activation function of Bas1p in an adenine-repressible step of ADE gene transcription.
Mol. Cell. Biol.
17:3272-3283[Abstract].
|
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
Present address: Laboratory of Biochemistry, Faculty of Food
Technology and Biotechnology, University of Zagreb, 10000 Zagreb, Croatia.
