Transcription activation at Class II CRP-dependent promoters: the role of different activating regions
Transcription activation at Class II CRP-dependent promoters: the role of different activating regions
Virgil A.
Rhodius
,
David M.
West
+
,
Christine L.
Webster
,
Stephen J. W.
Busby*
and
Nigel J.
Savery
School of Biochemistry, University of Birmingham, PO Box 363,
Birmingham
B15 2TT,
UK
Received September 27, 1996;
Revised and Accepted November 15, 1996
ABSTRACT
Transcription activation by the
Escherichia coli
cyclic AMP receptor protein (CRP) at Class II promoters is dependent on direct
interactions between two surface-exposed activating regions (AR1 and AR2) and two contact sites in RNA
polymerase. The effects on transcription activation of disrupting either AR1 or
AR2 have been measured at different Class II promoters. AR2 but not AR1 is essential for activation at all the Class II
promoters that were tested. The effects of single positive control
substitutions in AR1 and AR2 vary from one promoter to another: the effects of
the different substitutions are contingent on the -35 hexamer sequence. Abortive initiation assays have been used to
quantify the effects of positive control substitutions in each activating
region on the kinetics of transcription initiation at the Class II CRP- dependent promoter
pmelRcon
. At this promoter, the HL159 substitution in AR1 results in a defect in the
initial binding of RNA polymerase whilst the KE101 substitution in AR2 reduces
the rate of isomerization from the closed to the open complex.
INTRODUCTION
The
Escherichia coli
cyclic AMP (cAMP) receptor protein (CRP) is a global transcription factor
triggered by cAMP that activates gene expression at >100 different promoters (
1
). The cAMP-CRP complex binds as a dimer to 22 bp sequences at target promoters and
activates transcription by contacting RNA polymerase (RNAP) directly. Although
there is great diversity in the way that CRP-dependent promoters are organized, the most commonly found arrangement is
for transcription initiation to be dependent on a single CRP dimer centred at
or near position -41.5 (i.e. the 22 bp DNA site for CRP is centred between base pairs -41 and -42 upstream from the transcript start) (
2
). At such promoters, known as Class II promoters, the downstream subunit of the
CRP dimer overlaps the promoter -35 region.
In ternary transcriptionally competent CRP-RNAP-DNA complexes at Class II promoters, RNAP binds to promoter DNA
both upstream and downstream of the bound CRP: the C-terminal domain of the RNAP [alpha] subunit ([alpha]CTD) is responsible for the upstream contacts and `docks' to
the promoter DNA upstream of the bound CRP dimer (
3
,
4
). In these complexes, CRP makes contact with RNAP by two surface-exposed activating regions, defined by single amino acid positive control
substitutions that interfere with transcription activation at Class II
promoters. Activating region 1 (AR1) is a surface-exposed [beta]-turn (CRP residues 156-164) that is functional in the upstream subunit of the
CRP dimer and contacts [alpha]CTD, which is bound immediately upstream of CRP (
5
-
7
). Activating region 2 (AR2) is a positively charged region defined by
substitutions at H19, H21, E96 and K101, that is functional in the downstream
subunit of the CRP dimer and contacts a site in the N-terminal domain of the RNAP [alpha] subunit ([alpha]NTD) (
8
,
9
).
In this work, we report the effects of disrupting either AR1 or AR2 at a number
of different Class II promoters. We show that AR1 and AR2 play different roles
in transcription activation, that the requirement for AR1 and AR2 differs from
one promoter to another and that it is the -35 region sequence that sets the requirements for AR1 and AR2.
MATERIALS AND METHODS
Strains, plasmids and recombinant DNA methodology
The bacterial strains and plasmids used in this work are listed in Table
1
. Standard methods were employed for the isolation and manipulation of DNA
fragments. Different derivatives of pDCRP encoding mutant
crp
genes, listed in Table
1
, were made by site- directed mutagenesis.
The promoters used in this work are listed in Table
2
. All the promoters were cloned on
Eco
RI-
Hin
dIII fragments: manipulations were performed on fragments cloned in plasmid pAA121.
CC
(
-41.5
),
pmelR
,
pmelRcon
and
galP1
are all Class II CRP-dependent promoters with the CRP binding site centred at position -41.5 and have been described in our previous work: their base
sequences are shown in Figure
1
.
GalP1
[Delta]
4
is a derivative of
galP1
that is active in the absence of CRP but is repressed by cAMP-CRP. The VR1 derivative of
CC
(
-41.5
) was constructed using PCR, with a primer carrying the
pmelR
base sequence from -30 to -33. To make the VR2 derivative,
pmelR
upstream sequences were amplified by PCR, using primers that permitted the
product to be used as a primer to amplify
CC
(
-41.5
). This second round of PCR resulted in the
CC
(
-41.5
)
-VR2
derivative, carrying
pmelRcon
sequence upstream of the DNA site for CRP of
CC
(
-41.5
) (Fig.
1
and Table
2
). The
CW1
-
CW13
derivatives of
CC
(
-41.5
) were selected after PCR amplification of
CC
(
-41.5
) using a primer with random bases at promoter positions -30 to -33 (Fig.
1
and Table
5
).
Derivative of
E.coli melR
promoter with consensus CRP binding site centred at -41.5
Gaston
et al
. (12)
pmelRcon
Derivative of
pmelR
with point mutations at -45 and -49 that improve the CRP binding site
West
et al
. (18) (Fig. 1)
galP1
(p19T)
The
E.coli gal
regulatory region carrying a G -> T transversion at position -19 that inactivates
galP2
Attey
et al
. (3) (Fig. 1)
Promoter repressed by CRP
galP1
[Delta]
4
Derivative of
galP1
with a 4 bp deletion between the CRP binding site and the -10 hexamer such that it is repressed by CRP
Bell
et al
. (10)
Derivatives of
CC
(
-41.5
)
CC
(
-41.5
)
-VR1
Derivative of
CC
(
-41.5
) with bp -30 to -33 replaced with
pmelRcon
sequence
This work (Fig. 1)
CC
(
-41.5
)
-VR2
Derivative of
CC
(
-41.5
) with sequence upstream of CRP site replaced with
pmelRcon
upstream sequence
This work (Fig. 1)
CC
(
-41.5
)
-CW1-12
Derivative of
CC
(
-41.5
) with -35 hexamer CATGGA replaced with different CANNNN variants
This work (Table 5)
CC
(
-41.5
)
-CW13
Derivative of
CC
(
-41.5
) with -35 hexamer CATGGA replaced with TTGACA
This work (Table 5)
a
All promoters were cloned on
Eco
RI-
Hin
dIII fragments in pRW50.
Measurement of promoter activities
in vivo
To measure the activities of different promoters and their activation by
different CRP mutants,
Eco
RI-
Hin
dIII fragments carrying the promoters were cloned into the low copy number broad host range
lac
expression vector pRW50 and the recombinants were transformed into the [Delta]
crp
derivative of the [Delta]
lac
strain M182. CRP-dependent promoters are inactive in this strain but they are activated
when CRP or a derivative, cloned in plasmid pDCRP, is introduced by transformation. [beta]-Galactosidase activities of cells carrying different promoters were
measured as in our previous work (
10
,
11
) and are taken as a measure of CRP- dependent transcription activation.
Measurement of
pmelRcon
activity
in vitro
Abortive initiation was used to measure the kinetics of CRP- dependent transcription initiation as described previously (
12
,
13
), except that a Molecular Dynamics PhosphorImager and the program ImageQuant
3.3 was used for quantification of product formation. The
pmelRcon
promoter was carried on purified
Eco
RI-
Hin
dIII fragments, purified RNA polymerase holoenzyme was obtained from Amersham
International and wild-type CRP, Glu101 CRP or Leu159 CRP were purified as in our previous work (
4
). Turnover numbers (TON; see Table
4
) were determined from preformed complexes containing 50 nM RNA polymerase, 100
nM CRP and 2 nM template DNA. The complexes were incubated for 30 min at 37oC before the reaction was initiated by the addition of ApU and [[alpha]-
32
P]UTP. Samples were removed at 5 min intervals for 20 min and the rate of
product formation was determined.
One explanation is that AR1 is less important for activation at some promoters.
To confirm this explanation we needed to be sure that AR1 was completely
disrupted. To do this we used a synthetic oligodeoxynucleotide, degenerate for
codons 157-162 of CRP, to make a set of pDCRP derivatives with scrambled AR1
sequences. Three such mutant CRP derivatives, PDZ15, PDZ22 and PDZ26 (see Table
1
C), were picked at random and found to repress the
galP1
[Delta]
4
promoter normally. Table
3
shows activation of the
CC
(
-41.5
),
pmelR
,
pmelRcon
and
galP1
promoters with wild-type CRP or the HL159, TA158 , PDZ15, PDZ22 or PDZ26 mutants. The results
show that the activities of the scrambled CRP mutants are similar to the
activities of CRP carrying single substitutions in AR1. The sequence of AR1 in
the PDZ15, PDZ22 and PDZ26 mutants (Table
1
) is unrelated to the wild-type sequence and AR1, therefore, is probably completely inactive. We
conclude that the observed activation of
pmelR
,
pmelRcon
and
galP1
by these CRP mutants is because AR1 is not essential for CRP-dependent activation at these promoters (note, however, that AR1 must play
a role at these promoters, since substitutions in AR1 cause some reduction in
CRP-dependent activation).
AR2 of CRP is essential for transcription activation at Class II promoters
Substitutions in AR1 and AR2 have different kinetic effects on promoter activity
The above results show that AR1 and AR2 play different roles in transcription
initiation at Class II promoters. To investigate further the role of AR1 and
AR2 we determined the effects of the HL159 substitution (in AR1) and the KE101
substitution (in AR2) on the kinetics of CRP-dependent transcription activation at the
pmelRcon
promoter. We chose
pmelRcon
because it displays substantial CRP-dependent activity even with these single substitutions in AR1 or AR2
(Table
3
and
4
).
The process of transcription initiation at any promoter can be described by two
simple parameters, the binding constant of RNAP for the promoter (
K
B
) and the rate constant for isomerization from the closed to open complex (
k
f
), which can be measured simply using abortive initiation assays (
13
). Abortive assays were performed to measure the rate of production of the
product ApUpU at
pmelRcon
as a function of RNAP concentration in the presence of wild-type or mutant CRP. Figure
2
shows the resulting plots and Table
4
shows the kinetic parameters deduced from these plots. The data show that the
HL159 substitution (in AR1) primarily leads to a 10-fold reduction in the initial binding of RNAP to the promoter (
K
B
), whilst the KE101 substitution (in AR2) leads to a 10-fold reduction in the isomerization from closed to open complex (
k
f
). This shows that disruption of AR1 and AR2 affects two different steps in
transcription activation, at least at
pmelRcon
.
Substitutions in AR1 and AR2 have different effects at different promoters
The results in Table
3
show that the effects of single amino acid substitutions in AR1 or AR2 differ
from one Class II promoter to another. Some of the promoters are less sensitive
than others to disruptions in AR1 and AR2. The likely explanation is that some
promoters carry sequence elements that reduce the need for different activating
regions, whilst not relieving the requirement for CRP. To identify these
sequence elements, we constructed two hybrid promoters in which segments from
pmelRcon
were transplanted into
CC
(
-41.5
). In
CC
(
-41.5
)
-VR1
the sequence from positions -30 to -33 was substituted with
pmelRcon
sequence, replacing the last four positions of the -35 hexamer (Fig.
1
and Table
2
). In
CC
(
-41.5
)
-VR2
the sequence upstream of the DNA site for CRP in
CC
(
-41.5
) was replaced with
pmelRcon
sequence (Fig.
1
and Table
2
).
Figure
3
shows the activation of
CC
(
-41.5
),
pmelRcon
,
CC
(
-41.5
)
-VR1
and
CC
(
-41.5
)
-VR2
by wild-type CRP, CRP carrying the HL159 substitution in AR1, CRP carrying the
KE101 substitution in AR2 or CRP carrying both substitutions. At
CC
(
-41.5
) (Fig.
3
A) each of the single substitutions and the double substitution led to almost
total loss of CRP-dependent transcription activation. In contrast, at
pmelRcon
(Fig.
3
B), the single substitutions in AR1 or AR2 had less of an effect, as expected,
showing that the requirement for the two activating regions is reduced.
Interestingly, CRP carrying both HL159 and KE101 is completely unable to
activate
pmelRcon
.
Table 3
Percentage activation of four CRP-dependent promoters and percentage repression of the
galP1
[Delta]
4
promoter by different CRP AR1 and AR2 mutants (
in vivo
data)
CRP derivative
Activation (%)
Repression (%)
CC(-41.5)
pmelR
pmelRcon
galP1
galP1
[Delta]
4
Wild-type
100
100
100
100
100
AR1 mutants
TA158
6
65
21
68
100
HL159
5
74
25
76
100
PDZ15
2
26
25
67
101
PDZ22
3
20
27
54
101
PDZ26
3
41
20
67
99
AR2 mutants
HY19
8
8
24
67
107
KE101
2
7
16
45
102
HY19 KE101
0
1
1
0
100
The percentage activation and percentage repression were calculated from [beta]-galactosidase assays performed with M182[Delta]
crp
cells containing plasmid pDCRP, carrying the listed CRP derivatives, and plasmid pRW50, carrying the listed promoters
fused to the
lac
operon. [beta]-Galactosidase levels were measured in extracts of exponentially
growing cells and expressed in standard Miller units. Cells were grown in
Lennox broth supplemented with 25 [mu]g/ml tetracycline and 80 [mu]g/ml ampicillin. Each assay was performed at least three times from independent clones; data
differed by <10% from one experiment to another. The activities of the CRP-dependent promoters,
CC
(
-41.5
),
pmelR
,
pmelRcon
and
galP1
, in the presence and absence of wild-type CRP were designated 100 and 0% respectively. At
CC
(
-41.5
), 100% = 713 U and 0% = 18 U; at
pmelR
, 100% = 877 U and 0% = 19 U; at
pmelRcon
, 100% = 4437 U and 0% = 7 U; at
galP1
, 100% = 7214 U and 0% = 726 U. At
galP1
[Delta]
4
, the levels of repression in the presence and absence of wild-type CRP were designated 100 and 0% respectively: the activities of the
fully repressed and derepressed promoter were 23 and 510 U respectively.
.
In vitro
abortive initiation data for wild-type CRP, CRP HL159 and CRP KE101 at the
pmelRcon
promoter
CRP derivative
K
B
(M
-1
)
k
f
(s
-1
)
K
B
k
f
(M
-1
s
-1
)
Turnover no.
(product/promoter/min)
Wild-type CRP
5.21 * 10
7
(*1)
3.00 * 10
-2
(*1)
1.56 * 10
6
(*1)
76
CRP HL159
3.55 * 10
6
(*0.1)
5.77 * 10
-2
(*1.9)
2.04 * 10
5
(*0.13)
49
CRP KE101
1.89 * 10
8
(*3.6)
4.42 * 10
-3
(*0.1)
8.38 * 10
5
(*0.54)
71
The constants
k
f
and
K
B
k
f
were calculated respectively from the intercepts and slopes of the [tau] plots shown in Figure 2, using the program Microsoft Excel. The turnover
numbers (TON) were determined from preformed complexes as described in
Materials and Methods. These turnover numbers reflect differences in promoter
occupancy under our experimental conditions and do not correlate with promoter
activities
in vivo
(see Table 3). The numbers in parentheses express the effect of the mutants as
a factor of the value with wild-type CRP. The tight coupling of
pmelRcon
activity to CRP precludes the accurate measurement of
k
f
and
K
B
in the absence of CRP
The results show that the activity of
CC
(
-41.5
)
-VR1
(Fig.
3
C) is very similar to
pmelRcon
, whilst
CC
(
-41.5
)
-VR2
is very similar to
CC
(
-41.5
) (Fig.
3
D). This shows that, at least for
CC
(
-41.5
) and
pmelRcon
, the requirement for AR1 and AR2 is fixed by the base sequence from -30 to -33, rather than the sequence upstream of the DNA site for CRP. To
check this conclusion, upstream sequences from both
galP1
and the
E.coli rrnB P1
promoter were also cloned upstream of the DNA site for CRP in
CC
(
-41.5
): these sequences resulted in a 3- to 8-fold increase in CRP-dependent activity but did not affect the requirements for AR1
and AR2 (data not shown).
13 Busby,S., Kolb,A. and Minchin,S. (1994) In Kneale,G.G. (ed.), Methods in Molecular Biology, Vol. 30, DNA-Protein Interactions. Humana Press, NJ, pp. 397-412.
14 Casadaban,M. and Cohen,S. (1980) J. Mol. Biol., 138, 179-207.MEDLINE Abstract
15 Busby,S., Kotlarz,D. and Buc,H. (1983) J. Mol. Biol., 167, 259-274.MEDLINE Abstract
16 Kelsall,A., Evans,C. and Busby,S. (1985) FEBS Lett., 180, 155-159.
