JOURNAL OF BACrERIOLOGY, Nov. 1992, p. 6831-6839
Vol. 174, No. 21
0021-9193/92/216831-09$02.00/0 Copyright © 1992, American Society for Microbiology
Tripartite Structure of the Saccharomyces cerevisiae Arginase (CAR1) Gene Inducer-Responsive Upstream Activation Sequence MARINDA VIUJOEN,1 LADISLAU Z. KOVARI,2 IULIA A. KOVARI,2 HEUI-DONG PARK 2 HENDRIK J. J. vAN VUUREN,'1 AND TERRANCE G. COOPER2* Department of Microbiology and Institute for Biotechnology, University of Stellenbosch, Stellenbosch, South Africa,' and Department of Microbiology and Immunology, University of Tennessee, Memphis, Tennessee 381632 Received 3 August 1992/Accepted 27 August 1992
Arginase (CARI) gene expression in Saccharomyces cerevisiae is induced by arginine. The 5' regulatory region of CAR) contains four separable regulatory elements-two inducer-independent upstream activation sequences (UASs) (UASCl and UASC2), an inducer-dependent UAS (UASJ), and an upstream repression sequence (URSI) which negatively regulates CA4R and many other yeast genes. Here we demonstrate that three homologous DNA sequences originally reported to be present in the inducer-responsive UASI are in fact three exchangeable elements (UASI.A, UASI-B, and UASI.C). Although two of these elements, either the same or different ones, are required for transcriptional activation to occur, all three are required for maximal levels of induction. The elements operate in all orientations relative to one another and to the TATA sequence. All three UAS, elements bind protein(s); protein binding does not require arginine or overproduction of any of the putative arginine pathway regulatory proteins. The UASI-protein complex was also observed even when extracts were derived from arg80/argRI or arg81/argRll deletion mutants. Similar sequences situated upstream of ARG5,6 andARG3 and reported to negatively regulate their expression are able to functionally substitute for the C4RI UASJ elements and mediate reporter gene expression. CARI (arginase) gene expression in Saccharomyces cerevisiae has been studied as a model of biochemical mechanisms underlying eucaryotic transcription and the cis- and trans-acting factors by which it is regulated. CARI mRNA production is induced to a high level when arginine, the native inducer, is added to the culture medium (1, 5, 17-20, 27, 36). High-level CARI mRNA production is not observed, however, when a readily catabolized nitrogen source such as asparagine or glutamine is present (1, 3, 5, 18-20, 27); i.e., CARI expression exhibits apparent sensitivity to nitrogen catabolite repression (NCR) (1, 3, 5, 18-20, 27). CARI NCR sensitivity was generally accepted to result from a discrete transcriptional control process. However, it has recently been shown that CARI sensitivity to NCR occurs via NCRmediated inducer exclusion (4). Dissection of the CAR1 5'-flanking region to identify the sequences required for transcriptional activation and its regulation revealed four separate elements (11-13, 15, 22, 28-31). The first element identified was URS1, a repressorbinding site that functions not only in CARI but also in many other genes, e.g., those participating in meiosis, mating type control, heat shock response, carbon metabolism, inositol metabolism, and oxidative metabolism (2, 9, 10, 14, 15, 22, 28-31, 35, 37). The protein that binds to this site has been purified, and the genes encoding it have been cloned and sequenced (14, 14a). The CARI promoter also contains three discrete upstream activation sequences (UASs) (11-13, 28, 31, 33). The two most 5' UASs, UASCl and UASc2, operate constitutively (11-13). Consistent with its unregulated operation, UASCl has been shown to consist of ABF1 and RAP1 sites (12, 13). The UASc site and its associated proteins are *
currently under study. Thus far, three RAP1 sites and one ABF1 site have been identified (12, 13). However, additional evidence suggests that one or more sites remain to be identified (11). Operation of the third and weakest of the three C4RI UASs, UASI, is completely inducer dependent (Fig. 6 of reference 11). A model for the functional interaction of these sites and their associated proteins has been proposed (11-15, 23, 38). According to this model, the negative action of the strong URSI element balances the positive action of the strong, constitutively functioning UASCl and UASc2 elements (1115). The balance is then tipped in the direction of expression when arginine is present, or quiescence when it is not, by operation of the inducer-dependent UASJ element (11-15). Observations by Kovari et al. demonstrated that three separable elements probably participate in induced expression mediated by UAS, (11, 34). These experiments localized the three elements to three homologous sequences (nucleotides -233 to -223, -209 to -199, and -170 to -160 [11]). However, little more is known about these putative elements, the proteins required for their operation, and whether they are functionally related. Therefore, we investigated the structure of UAS, and its component parts in greater detail. We found that (i) UASJ consists of three 27-bp interchangeable elements that operate with different efficacies, (ii) only two of the three homologous elements are required at once for minimal UAS function, (iii) these elements operate in orientation-independent fashion both with respect to one another and with respect to the TATA sequence, and (iv) a protein specifically binds to the UAS, elements reported here and possesses characteristics quite different from those of the protein reported earlier (16). (These results have been presented in part earlier [34].)
Corresponding author. 6831
6832
VILJOEN ET AL.
J. -245
LK207 (UASI-A)
-219
AGTCTCTAGCTCTTGCCCTTCGCAAAG -213
LK209 (UASI.B)
-187
TGCTGCTAATGGCAATCAACAGCGCAT -182
-156
LK157
(UASI.C)
LK211
(UASI.Cmt) GCTCCgaGAATTTTagACggAGCGGTA
LK212
(ARG3)
GCTCGCTGAATTTTTCACTTAGCGGTA -182
-
-156
6g1
-817
ACAAGGAATAAACTGCCTTTAGAGGTG -126
LK213 (ARG5,6)
-105
TTGTTCGCTATCCATTTCCATTAGG -245
pMV165
BACTERIOL.
B
A
'iGTCTCTAGCTCTTGCCCTTCGCAAAGCACCG'TGCTGCTAATGGCAAT C_---__
-156
CAACAGCGCATCGCC'GCTCGCTGAATTTTTCACTTAGCGGTA' pMV166
A mt-
-*-
B
'AGTCTCTAGCTCTTGgCCTTCGCAAAG CACC*IGCTGCTAATGGCAAT c
CAACAGCGCATtGCdGCTCGCTGAATTTTTCACTTAGCGGTA -245
pMV167
A--)
Bmt
AGTCTCTAGCTCTTGCCCTTCGCAAAGCACCGTGCTGCTAATGcCAAT .~~~~~
-156
CAACAGCGCATC'GCCG'TCGCTGAATTTTTCACTTAGCGGTA' PMV168
A -
r
-galactosidase AMG gLU
UASI Pasmid
pMV165
In StructuM - WT b || b
WT
I-a9---8I|| --WT-||
pMV167
1wT - I4 - --
pMV168
-|-f' 1~ a9 -- II *--
pMV169 pHP41
No Inserl
-WT
WT
g9
12
730
60.8
II
10
392
39.2
I
ii
60
5.5
5
35
3.9
5
6
1.2
30
24
0.8
-WT
pMV166
-gI
Il
9
L-omA
ndedn
FIG. 3. Reporter gene (lacZ) expression supported by synthetic DNA fragments containing wild-type (WT) and mutant UASJ elements from CARI. Plasmid pMV165 contains the entire wild-type element. Plasmids pMV166, pMV167, and pMV168 contain single point mutations (C-to-G transversion, indicated by g) in the UASI-A, UASI B, and UASI-C elements. The precise sequences of these constructions are shown in Fig. 1. The remaining elements in these constructions are wild type. Assay procedures are described in Materials and Methods. Fold induction is derived by dividing the 3-galactosidase activity observed with cells grown in arginine medium (ARG) by that observed with cells grown in glutamate medium (GLU). Arrows indicate the orientations of the native elements relative to one another. The native orientation and structure of UAS, as it appears upstream of CARI are shown at the top.
INDUCER-RESPONSIVE UAS OF THE YEAST C4RI GENE
VOL. 174, 1992
o
*n
es
_
0
EL
XL
0
Xu0 x o
u
la
Cs N
N
_
UASI-e Mutant Comp.
r
o
._0
UASI.C Comp.
UASI-r1 Mutant Comp.
UASI-C Comp.
UAS,.c Mutant Comp.
UASI-C Comp.
6835
0
O
CM
LI
rCM X
_
O
X
Q
C,
o
U
_
N
I
.U .i
.. . ..
twootoew--
w bt
W4
JL-
.-.&
NW U.
"*-A
t. X##::
?"
*4
i:,
*.A
kB.
:,
.4
.L.A- Ak;;A id:M N::ih!* a --
.CA i,'.4
". ow n ". S.4
V....v
IW.
4w
W.-
4' ."-'i k''
,-
.1~
Sil__|~~~~~~~~~~~~~~~~~~~~~~~~~ q[""
__
-
:;-
.:.
..:
:z,
i.
4.
A
B
C
E
F
H
J
PROBE:
G
UASIC
K
L
M
N
O
A
4 i.11 41M
B
C
D
E
F
G
H
J
K
L
M
N
O
PROBE: UASlg
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
PROBE: UASI.A
FIG. 5. Competition between the CARI UASI-A, UASI and UASI-C elements for protein binding. 32P-labeled oligonucleotides were used indicated below the panels. Cell extract was omitted from the reaction mixture in lanes H (-EXT). The reaction mixtures in lanes F through A contained increasing amounts (micrograms) of the unlabeled CARI UASI-C oligonucleotide as the competitor. The reaction mixtures in lanes J through 0 contained increasing amounts (micrograms) of the unlabeled CARl UASI-C mutant oligonucleotide. Specific competitor was omitted from reaction mixtures shown in lanes G and I (-COMP). All reaction mixtures contained a 200-fold excess of poly(dI-dC) as a nonspecific competitor. After a 20-min incubation at room temperature, complexes were resolved on a 4% nondenaturing polyacrylamide gel. The nucleotide sequences of probes and competitors are shown in Fig. 1. The arrow indicates a protein whose binding to the UASI- fragment was successfully competed by the UASI- fragment. B,
as probes as
the predicted results
were
observed experimentally. Two
UASI-C sequences (plasmid pMV147) supported the greatest amount of reporter gene expression, followed by UASIB plus UASI-C, UASI-B plus UASI and UASI-A plus UASI-c (plasmids pMV143, pMV146, and pMV144, respectively). The response to an inducer, i.e., the fold induction, roughly followed the induced level of reporter gene expression. An interesting observation that is too close to experimental error to be taken seriously without further investigation is the increase in the amount of 3-galactosidase production supported by a plasmid containing two copies of the UASI-C sequence (plasmid pMV147) in cells that were not provided with inducer (Glu in Fig. 4). Note that plasmids containing elements UASI-B plus UASI-A and UASI-A plus UASI (plasmids pMV145 and pMV148) supported reporter gene expression that was below the level of the parent vector (plasmid pHP41). Protein binding to the UASI_A, UASI B, and UASJEC elements. The demonstration above that all three UAS, elements function in vivo predicted that they should bind one or more proteins. To test this prediction, we assessed the ability of three fragments containing only the UASIA, UASI B, or UASI-C sequence (Fig. 1) to bind proteins in an EMSA. We used a crude extract as the source of protein for this experiment to avoid the possibility of losing proteins that might bind to the UAS, elements. Protein binding to each of the three labeled fragments was competed by either the UASI-C fragment or a mutant form of this fragment (Fig. 1). As shown in Fig. 5, multiple bands were observed and the patterns of DNA-protein complex mobilities were very similar regardless of which DNA fragment was labeled. Note that the autoradiographs in the three panels of this figure were exposed for different lengths of time. The UASI-C fragment successfully competed its own binding to one of these proteins (Fig. 5, left panel, arrow, lanes A to H). In B,
A
UASI-C fragment was not an effective competitor (Fig. 5, left panel, arrow, lanes H to 0). The same experiment was repeated with labeled UASI-B and UASIA DNA fragments (Fig. 5, center and right panels, respectively, arrow, lanes A to H). Again the wild-type UASIc, but not the mutant fragment, was an effective competitor (Fig. 5, center and right panels, arrow, lanes H to 0). These data indicated that all three DNA fragments exhibited very similar profiles of protein binding and had at least one specific, competable protein-binding site in com-
contrast, the mutant
mon.
The specifically competable DNA-protein complexes observed in this experiment were different from the one reported by Messenguy et al., which could not be observed in protein preparations derived from an arg8l/argR!I disruption mutant (16). The protein-DNA complexes we observed were present when extracts from argRI and argRII deletion mutants were used as the source of protein (Fig. 6, lanes A to I, complex 6). Hence, they were not dependent on either ARGRI or ARGRII products. Lack of orientation specificity of the UASI_A, UASI-B, and UASI C sequences. Portions of the UASIB and UASI-C sequences, respectively (see Fig. lOB), could be folded into a hairpin loop (11, 16). However, the fact that three rather than two sequences participated in UASI-mediated inducible CARI expression led us to suspect that it was unlikely for a hairpin loop to function in transcriptional activation mediated by UAS,. However, if a hairpin loop were absolutely required for transcriptional activation, then a similar absolute orientation specificity would exist because the elements were so close together. We tested the orientation requirements by placing the UASIB and UASI-C elements in all possible orientations with respect to one another. As shown in Fig. 7, all of them were functional to various degrees ranging from 20 to approximately 100% efficiency.
6836
VILJOEN ET AL.
PROBES:
-
Vol. 174, No. 21
0021-9193/92/216831-09$02.00/0 Copyright © 1992, American Society for Microbiology
Tripartite Structure of the Saccharomyces cerevisiae Arginase (CAR1) Gene Inducer-Responsive Upstream Activation Sequence MARINDA VIUJOEN,1 LADISLAU Z. KOVARI,2 IULIA A. KOVARI,2 HEUI-DONG PARK 2 HENDRIK J. J. vAN VUUREN,'1 AND TERRANCE G. COOPER2* Department of Microbiology and Institute for Biotechnology, University of Stellenbosch, Stellenbosch, South Africa,' and Department of Microbiology and Immunology, University of Tennessee, Memphis, Tennessee 381632 Received 3 August 1992/Accepted 27 August 1992
Arginase (CARI) gene expression in Saccharomyces cerevisiae is induced by arginine. The 5' regulatory region of CAR) contains four separable regulatory elements-two inducer-independent upstream activation sequences (UASs) (UASCl and UASC2), an inducer-dependent UAS (UASJ), and an upstream repression sequence (URSI) which negatively regulates CA4R and many other yeast genes. Here we demonstrate that three homologous DNA sequences originally reported to be present in the inducer-responsive UASI are in fact three exchangeable elements (UASI.A, UASI-B, and UASI.C). Although two of these elements, either the same or different ones, are required for transcriptional activation to occur, all three are required for maximal levels of induction. The elements operate in all orientations relative to one another and to the TATA sequence. All three UAS, elements bind protein(s); protein binding does not require arginine or overproduction of any of the putative arginine pathway regulatory proteins. The UASI-protein complex was also observed even when extracts were derived from arg80/argRI or arg81/argRll deletion mutants. Similar sequences situated upstream of ARG5,6 andARG3 and reported to negatively regulate their expression are able to functionally substitute for the C4RI UASJ elements and mediate reporter gene expression. CARI (arginase) gene expression in Saccharomyces cerevisiae has been studied as a model of biochemical mechanisms underlying eucaryotic transcription and the cis- and trans-acting factors by which it is regulated. CARI mRNA production is induced to a high level when arginine, the native inducer, is added to the culture medium (1, 5, 17-20, 27, 36). High-level CARI mRNA production is not observed, however, when a readily catabolized nitrogen source such as asparagine or glutamine is present (1, 3, 5, 18-20, 27); i.e., CARI expression exhibits apparent sensitivity to nitrogen catabolite repression (NCR) (1, 3, 5, 18-20, 27). CARI NCR sensitivity was generally accepted to result from a discrete transcriptional control process. However, it has recently been shown that CARI sensitivity to NCR occurs via NCRmediated inducer exclusion (4). Dissection of the CAR1 5'-flanking region to identify the sequences required for transcriptional activation and its regulation revealed four separate elements (11-13, 15, 22, 28-31). The first element identified was URS1, a repressorbinding site that functions not only in CARI but also in many other genes, e.g., those participating in meiosis, mating type control, heat shock response, carbon metabolism, inositol metabolism, and oxidative metabolism (2, 9, 10, 14, 15, 22, 28-31, 35, 37). The protein that binds to this site has been purified, and the genes encoding it have been cloned and sequenced (14, 14a). The CARI promoter also contains three discrete upstream activation sequences (UASs) (11-13, 28, 31, 33). The two most 5' UASs, UASCl and UASc2, operate constitutively (11-13). Consistent with its unregulated operation, UASCl has been shown to consist of ABF1 and RAP1 sites (12, 13). The UASc site and its associated proteins are *
currently under study. Thus far, three RAP1 sites and one ABF1 site have been identified (12, 13). However, additional evidence suggests that one or more sites remain to be identified (11). Operation of the third and weakest of the three C4RI UASs, UASI, is completely inducer dependent (Fig. 6 of reference 11). A model for the functional interaction of these sites and their associated proteins has been proposed (11-15, 23, 38). According to this model, the negative action of the strong URSI element balances the positive action of the strong, constitutively functioning UASCl and UASc2 elements (1115). The balance is then tipped in the direction of expression when arginine is present, or quiescence when it is not, by operation of the inducer-dependent UASJ element (11-15). Observations by Kovari et al. demonstrated that three separable elements probably participate in induced expression mediated by UAS, (11, 34). These experiments localized the three elements to three homologous sequences (nucleotides -233 to -223, -209 to -199, and -170 to -160 [11]). However, little more is known about these putative elements, the proteins required for their operation, and whether they are functionally related. Therefore, we investigated the structure of UAS, and its component parts in greater detail. We found that (i) UASJ consists of three 27-bp interchangeable elements that operate with different efficacies, (ii) only two of the three homologous elements are required at once for minimal UAS function, (iii) these elements operate in orientation-independent fashion both with respect to one another and with respect to the TATA sequence, and (iv) a protein specifically binds to the UAS, elements reported here and possesses characteristics quite different from those of the protein reported earlier (16). (These results have been presented in part earlier [34].)
Corresponding author. 6831
6832
VILJOEN ET AL.
J. -245
LK207 (UASI-A)
-219
AGTCTCTAGCTCTTGCCCTTCGCAAAG -213
LK209 (UASI.B)
-187
TGCTGCTAATGGCAATCAACAGCGCAT -182
-156
LK157
(UASI.C)
LK211
(UASI.Cmt) GCTCCgaGAATTTTagACggAGCGGTA
LK212
(ARG3)
GCTCGCTGAATTTTTCACTTAGCGGTA -182
-
-156
6g1
-817
ACAAGGAATAAACTGCCTTTAGAGGTG -126
LK213 (ARG5,6)
-105
TTGTTCGCTATCCATTTCCATTAGG -245
pMV165
BACTERIOL.
B
A
'iGTCTCTAGCTCTTGCCCTTCGCAAAGCACCG'TGCTGCTAATGGCAAT C_---__
-156
CAACAGCGCATCGCC'GCTCGCTGAATTTTTCACTTAGCGGTA' pMV166
A mt-
-*-
B
'AGTCTCTAGCTCTTGgCCTTCGCAAAG CACC*IGCTGCTAATGGCAAT c
CAACAGCGCATtGCdGCTCGCTGAATTTTTCACTTAGCGGTA -245
pMV167
A--)
Bmt
AGTCTCTAGCTCTTGCCCTTCGCAAAGCACCGTGCTGCTAATGcCAAT .~~~~~
-156
CAACAGCGCATC'GCCG'TCGCTGAATTTTTCACTTAGCGGTA' PMV168
A -
r
-galactosidase AMG gLU
UASI Pasmid
pMV165
In StructuM - WT b || b
WT
I-a9---8I|| --WT-||
pMV167
1wT - I4 - --
pMV168
-|-f' 1~ a9 -- II *--
pMV169 pHP41
No Inserl
-WT
WT
g9
12
730
60.8
II
10
392
39.2
I
ii
60
5.5
5
35
3.9
5
6
1.2
30
24
0.8
-WT
pMV166
-gI
Il
9
L-omA
ndedn
FIG. 3. Reporter gene (lacZ) expression supported by synthetic DNA fragments containing wild-type (WT) and mutant UASJ elements from CARI. Plasmid pMV165 contains the entire wild-type element. Plasmids pMV166, pMV167, and pMV168 contain single point mutations (C-to-G transversion, indicated by g) in the UASI-A, UASI B, and UASI-C elements. The precise sequences of these constructions are shown in Fig. 1. The remaining elements in these constructions are wild type. Assay procedures are described in Materials and Methods. Fold induction is derived by dividing the 3-galactosidase activity observed with cells grown in arginine medium (ARG) by that observed with cells grown in glutamate medium (GLU). Arrows indicate the orientations of the native elements relative to one another. The native orientation and structure of UAS, as it appears upstream of CARI are shown at the top.
INDUCER-RESPONSIVE UAS OF THE YEAST C4RI GENE
VOL. 174, 1992
o
*n
es
_
0
EL
XL
0
Xu0 x o
u
la
Cs N
N
_
UASI-e Mutant Comp.
r
o
._0
UASI.C Comp.
UASI-r1 Mutant Comp.
UASI-C Comp.
UAS,.c Mutant Comp.
UASI-C Comp.
6835
0
O
CM
LI
rCM X
_
O
X
Q
C,
o
U
_
N
I
.U .i
.. . ..
twootoew--
w bt
W4
JL-
.-.&
NW U.
"*-A
t. X##::
?"
*4
i:,
*.A
kB.
:,
.4
.L.A- Ak;;A id:M N::ih!* a --
.CA i,'.4
". ow n ". S.4
V....v
IW.
4w
W.-
4' ."-'i k''
,-
.1~
Sil__|~~~~~~~~~~~~~~~~~~~~~~~~~ q[""
__
-
:;-
.:.
..:
:z,
i.
4.
A
B
C
E
F
H
J
PROBE:
G
UASIC
K
L
M
N
O
A
4 i.11 41M
B
C
D
E
F
G
H
J
K
L
M
N
O
PROBE: UASlg
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
PROBE: UASI.A
FIG. 5. Competition between the CARI UASI-A, UASI and UASI-C elements for protein binding. 32P-labeled oligonucleotides were used indicated below the panels. Cell extract was omitted from the reaction mixture in lanes H (-EXT). The reaction mixtures in lanes F through A contained increasing amounts (micrograms) of the unlabeled CARI UASI-C oligonucleotide as the competitor. The reaction mixtures in lanes J through 0 contained increasing amounts (micrograms) of the unlabeled CARl UASI-C mutant oligonucleotide. Specific competitor was omitted from reaction mixtures shown in lanes G and I (-COMP). All reaction mixtures contained a 200-fold excess of poly(dI-dC) as a nonspecific competitor. After a 20-min incubation at room temperature, complexes were resolved on a 4% nondenaturing polyacrylamide gel. The nucleotide sequences of probes and competitors are shown in Fig. 1. The arrow indicates a protein whose binding to the UASI- fragment was successfully competed by the UASI- fragment. B,
as probes as
the predicted results
were
observed experimentally. Two
UASI-C sequences (plasmid pMV147) supported the greatest amount of reporter gene expression, followed by UASIB plus UASI-C, UASI-B plus UASI and UASI-A plus UASI-c (plasmids pMV143, pMV146, and pMV144, respectively). The response to an inducer, i.e., the fold induction, roughly followed the induced level of reporter gene expression. An interesting observation that is too close to experimental error to be taken seriously without further investigation is the increase in the amount of 3-galactosidase production supported by a plasmid containing two copies of the UASI-C sequence (plasmid pMV147) in cells that were not provided with inducer (Glu in Fig. 4). Note that plasmids containing elements UASI-B plus UASI-A and UASI-A plus UASI (plasmids pMV145 and pMV148) supported reporter gene expression that was below the level of the parent vector (plasmid pHP41). Protein binding to the UASI_A, UASI B, and UASJEC elements. The demonstration above that all three UAS, elements function in vivo predicted that they should bind one or more proteins. To test this prediction, we assessed the ability of three fragments containing only the UASIA, UASI B, or UASI-C sequence (Fig. 1) to bind proteins in an EMSA. We used a crude extract as the source of protein for this experiment to avoid the possibility of losing proteins that might bind to the UAS, elements. Protein binding to each of the three labeled fragments was competed by either the UASI-C fragment or a mutant form of this fragment (Fig. 1). As shown in Fig. 5, multiple bands were observed and the patterns of DNA-protein complex mobilities were very similar regardless of which DNA fragment was labeled. Note that the autoradiographs in the three panels of this figure were exposed for different lengths of time. The UASI-C fragment successfully competed its own binding to one of these proteins (Fig. 5, left panel, arrow, lanes A to H). In B,
A
UASI-C fragment was not an effective competitor (Fig. 5, left panel, arrow, lanes H to 0). The same experiment was repeated with labeled UASI-B and UASIA DNA fragments (Fig. 5, center and right panels, respectively, arrow, lanes A to H). Again the wild-type UASIc, but not the mutant fragment, was an effective competitor (Fig. 5, center and right panels, arrow, lanes H to 0). These data indicated that all three DNA fragments exhibited very similar profiles of protein binding and had at least one specific, competable protein-binding site in com-
contrast, the mutant
mon.
The specifically competable DNA-protein complexes observed in this experiment were different from the one reported by Messenguy et al., which could not be observed in protein preparations derived from an arg8l/argR!I disruption mutant (16). The protein-DNA complexes we observed were present when extracts from argRI and argRII deletion mutants were used as the source of protein (Fig. 6, lanes A to I, complex 6). Hence, they were not dependent on either ARGRI or ARGRII products. Lack of orientation specificity of the UASI_A, UASI-B, and UASI C sequences. Portions of the UASIB and UASI-C sequences, respectively (see Fig. lOB), could be folded into a hairpin loop (11, 16). However, the fact that three rather than two sequences participated in UASI-mediated inducible CARI expression led us to suspect that it was unlikely for a hairpin loop to function in transcriptional activation mediated by UAS,. However, if a hairpin loop were absolutely required for transcriptional activation, then a similar absolute orientation specificity would exist because the elements were so close together. We tested the orientation requirements by placing the UASIB and UASI-C elements in all possible orientations with respect to one another. As shown in Fig. 7, all of them were functional to various degrees ranging from 20 to approximately 100% efficiency.
6836
VILJOEN ET AL.
PROBES:
-
