doi:10.1006/jmbi.2000.3737 available online at http://www.idealibrary.com on
J. Mol. Biol. (2000) 299, 311±324
Interactions between Activating Region 3 of the
Escherichia coli Cyclic AMP Receptor Protein and
Region 4 of the RNA Polymerase s 70 Subunit:
Application of Suppression Genetics
Virgil A. Rhodius and Stephen J. W. Busby*
School of Biosciences, The
University of Birmingham
Birmingham B15 2TT, UK
The Escherichia coli cyclic AMP receptor protein, CRP, induces transcription at Class II CRP-dependent promoters by making three different activatory contacts with different surfaces of holo RNA polymerase. One
contact surface of CRP, known as Activating Region 3 (AR3), is functional in the downstream subunit of the CRP dimer and is predicted to
interact with region 4 of the RNAP s70 subunit. We have previously
shown that a mutant CRP derivative that activates transcription primarily
via AR3, CRP HL159 KE101 KN52, requires the positively charged residues K593, K597 and R599 in s70 for activation. Here, we have used the
positive control substitution, EK58, to disrupt AR3-dependent activation
by CRP HL159 KE101 KN52. We then screened random mutant libraries
and an alanine scan library of s70 for candidates that restore activation
by CRP HL159 KE101 KN52 EK58. We found that changes at R596 and
R599 in s70 can restore activation by CRP HL159 KE101 KN52 EK58.
This suggests that the side-chains of both R596 and R599 in s70 clash
with K58 in CRP. Maximal activation by CRP HL159 KE101 KN52 EK58
is achieved with the substitutions RE596 or RD596 in s70. We propose
that there are speci®c charge-charge interactions between E596 or D596
in s70 and K58 in AR3. Thus, no increase in activation is observed in the
presence of another positive control substitution, EG58 (CRP HL159
KE101 KN52 EG58). Similarly, both s70 RE596 and s70 RD596 can restore
activation by CRP EK58 but not CRP EG58, and they both decrease activation by wild-type CRP. We suggest that E596 and D596 in s70 can
positively interact with K58 in AR3, thereby enhancing activation, but
negatively interact with E58, thereby decreasing activation. The substitution, KA52 in AR3 increases Class II CRP-dependent activation by
removing an inhibitory lysine residue. However, this increase is not
observed in the presence of either s70 RE596 or s70 RD596. We conclude
that the inhibitory side-chain, K52 in AR3, clashes with R596 in s70.
Finally, we show that the s70 RE596 and RD596 substitutions affect CRPdependent activation from Class II, but not Class I, promoters.
# 2000 Academic Press
*Corresponding author
Keywords: Escherichia coli; cyclic AMP receptor protein; CRP; RNA
polymerase s70 subunit; transcription activation
Introduction
Abbreviations used: CRP, cyclic AMP receptor
protein; RNAP, holo form of the Escherichia coli DNAdependent RNA polymerase; AR1, Activating Region 1;
AR2, Activating Region 2; AR3, Activating Region 3;
aCTD, C-terminal domain of the RNAP; aNTD,
N-terminal domain of the RNAP.
E-mail address of the corresponding author:
s.j.w.busby@bham.ac.uk
0022-2836/00/020311±14 $35.00/0
Transcription initiation in prokaryotic cells
involves the multisubunit RNA polymerase core
enzyme and one of several different types of sigma
factor. In Escherichia coli, s70 (encoded by rpoD) is
the primary ``housekeeping'' sigma that is responsible for the expression of the majority of genes
during logarithmic growth. The s70 subunit of
# 2000 Academic Press
312
RNA polymerase holoenzyme (RNAP) recognises
and binds to two sequence elements within promoters called the ÿ10 and ÿ35 hexamers (reviewed
by Gross et al., 1998). It has been proposed that the
ÿ10 hexamer is recognised by a single a-helix
within conserved region 2.4 of s70 and the ÿ35
hexamer is recognised by a helix-turn-helix motif
located in conserved region 4.2 near the C terminus
of s70. An additional function of the s70 subunit of
RNAP is to serve as a target for several transcription factors that bind to DNA sites that overlap the
ÿ35 segment of promoters. Target sites on s70
have been identi®ed by single amino acid substitutions that speci®cally disrupt the function of certain activators but do not affect basal transcription
levels. Many activators appear to contact a target
adjacent to the helix-turn-helix-motif in region 4.2
(Figure 1; reviewed by Gross et al., 1998 and
Rhodius & Busby, 1998).
The E. coli cyclic AMP receptor protein (CRP;
also known as the catabolite gene activator protein,
CAP) is a dimeric transcription factor activated by
the binding of cyclic AMP (cAMP). At CRPdependent promoters, CRP activates transcription
by making direct protein-protein contacts with
RNAP (reviewed by Busby & Ebright, 1999). Class
I CRP-dependent promoters contain a single CRPbinding site located upstream of the binding site
for RNAP. At these promoters, CRP activates transcription by contacting the C-terminal domain of
the RNAP a-subunit (aCTD) via a surface exposed
patch known as Activating Region 1 (AR1) in the
downstream subunit of the bound CRP dimer.
Class II CRP-dependent promoters contain a single
CRP-binding site which overlaps the ÿ35 hexamer.
At such promoters, CRP functions as an ``ambidextrous'' activator by making multiple contacts with
RNAP (Busby & Ebright, 1997). AR1 in the
upstream subunit of the bound CRP dimer interacts with aCTD of RNAP and Activating Region 2
(AR2) in the downstream subunit interacts with
the N-terminal domain of the a-subunit (aNTD).
CRP also contains a third separate activating
region known as Activating Region 3 (AR3), which
is functional in the downstream subunit of CRP
(Bell et al., 1990; Williams et al., 1991, 1996;
Rhodius & Busby, 2000). AR3 is composed of an
activatory determinant consisting of the negatively
charged residues D53, E54, E55 and E58, and an
inhibitory determinant consisting of the positively
charged residue, K52 (Rhodius & Busby, 2000).
Several lines of evidence argue that, at Class II
promoters, the target of AR3 is region 4 of the
RNAP s70 subunit. First, molecular modelling of
the CRP-DNA complex reveals that AR3 must be
in close proximity to the ÿ35 element, and consequently is ideally situated to interact with region 4
of s70. Second, Jin et al. (1995) showed that AR3 of
CRP could be cross-linked to s70 at Class II, but
not at Class I, promoters. Third, we demonstrated
that a CRP derivative functioning primarily via
AR3 requires the positively charged residues K593,
K597 and R599 in s70 for activation (Figure 1;
70 Suppressor Mutants
Figure 1. Schematic map of s70 illustrating conserved
regions and activator contact sites. The top part of the
Figure shows a linear representation of E. coli s70 illustrating the location of the four highly conserved regions
of the s70 family. Amino acid positions are indicated
below the s70 peptide. The bottom part of the
Figure shows residues 530 to 613 of s70 and conserved
regions 3.2, 4.1 and 4.2; amino acid positions are indicated above the partial s70 peptide. The position of the
predicted helix-turn-helix motif within region 4.2 is indicated and the sites of amino acid substitutions that
modify activation by different E. coli transcription
factors are shown: negatively charged residues on s70,
(red); positively charged residues, (blue); and all other
residues, (grey). Data for lcI are from Kuldell &
Hochschild (1994) and Li et al. (1994); for Ada, from
Landini et al. (1998) and Landini & Busby (1999); for the
CRP mutant, CRP HL159 KE101 KN52 and FNR, from
Lonetto et al. (1998); for PhoB, from Kim et al. (1995);
and for AraC, from Hu & Gross (1985).
Lonetto et al., 1998). These residues, located
immediately downstream of the helix-turn-helix
motif in region 4.2 of s70, form a positively
charged patch that complements the negatively
charged activatory determinant in AR3. However,
there is no direct evidence for interactions between
AR3 and region 4 of s70. Thus, to demonstrate an
interaction, we have identi®ed s70 mutants carrying changes that compensate for substitutions in
CRP that inactivate AR3. These s70 mutants contained substitutions in the positively charged
surface previously identi®ed by Lonetto et al.
(1998).
Results and Discussion
s 70 mutants that increase activation by CRP
HL159 KE101 KN52 EK58
We have used suppression genetics to identify
residues in s70 that are in close proximity to
313
70 Suppressor Mutants
residues in AR3 of CRP: thus, we screened a random mutant library of rpoD for candidates that
increased activation by a CRP mutant containing a
single positive control substitution in AR3. We
started with the CRP mutant, CRP HL159 KE101
KN52, which activates transcription primarily via
AR3 at the Class II CRP-dependent promoter,
pmelRcon (Lonetto et al., 1998; Rhodius & Busby,
2000). This mutant CRP contains the positive control substitutions HL159 and KE101 that disrupt
AR1 and AR2, respectively, and the substitution
KN52 that improves AR3. We have demonstrated
that E58 is the single most important side-chain
required for AR3-dependent activation (Rhodius &
Busby, 2000). Thus, we employed the charge reversal substitution, EK58, to create the mutant, CRP
HL159 KE101 KN52 EK58, which is unable to
activate pmelRcon. We then screened a plasmidencoded random mutant library of s70 for suppressors that increased activation by CRP HL159
KE101 KN52 EK58 at pmelRcon.
A library of rpoD mutants, in which codons 530
to 613 had been subjected to random mutagenesis,
was created using error-prone PCR (see Materials
and Methods). The rpoD mutants were carried in
the vector pVRs. To ensure similar levels of
expression in the presence of different s70 mutants
in pVRs, the rpoD gene is expressed from the constitutive ``extended ÿ10`` promoter galP1-27 (Busby
et al., 1987). Since transcription initiation at
extended ÿ10 promoters does not require region 4
of s70 (Chan & Busby, 1989; Kumar et al., 1993),
we reasoned that s70 mutants with altered ÿ35
promoter recognition would not alter galP1-27
activity, and hence their own level of expression.
We then constructed a screening strain that we
could use to identify s70 mutants that increased
activation by CRP HL159 KE101 KN52 EK58 at
pmelRcon. First, we transformed the crp lac
strain, M182crp, with the plasmid, pRW50 carrying the CRP-dependent promoter fusion, pmelRcon::lacZ. We then transformed the low copy
number plasmid, pLG339CRP, encoding CRP
HL159 KE101 KN52 EK58. This gives a Lacÿ
phenotype since this CRP mutant is unable to activate pmelRcon. Next, we introduced the plasmid,
pVRs, encoding the rpoD random mutant library.
To identify s70 mutants that restored activation by
CRP HL159 KE101 KN52 EK58 at pmelRcon, transformants were plated on MacConkey Lactose agar
indicator plates and screened for Lac‡ colonies. We
reasoned that only s70 mutants that increased activation of pmelRcon by CRP HL159 KE101 KN52
EK58 and, therefore, were trans dominant to
chromosomal expressed rpoD, would result in a
Lac‡ phenotype. All other s70 mutants would
result in a Lacÿ phenotype.
A total of 22,000 transformants were screened
and six mutants were isolated that scored Lac‡,
each from independent PCR libraries. For each
mutant, the pVRs plasmid was isolated and the
cloned rpoD gene sequenced and found to encode
for mutant s70 with a Gly, Ser or Cys residue at
position 596 (Table 1A). Since PCR-based mutagenesis can only create a limited set of amino acid substitutions, a second rpoD library was constructed,
in which codon 596 was completely randomised
(see Materials and Methods). This library was
screened in the same way as above, and seven
mutants that scored Lac‡ were selected from 700
transformants. As before, the plasmid DNA, pVRs
was isolated and the cloned rpoD gene sequenced.
The pVRs mutants were found to encode s70 with
a Gly, Ser or Cys substitution at position 596, as
before, but also with a Glu or Asp substitution
(Table 1B).
Activity of the s 70 suppressor mutants
The effects of the different s70 mutants on activation by CRP HL159 KE101 KN52 EK58 at pmelRcon were quanti®ed in vivo using b-galactosidase
assays. The pVRs plasmids encoding the different
s70 mutants, were transformed into M182crp cells
carrying pRW50/pmelRcon::lacZ and pLG339CRP,
encoding wild-type CRP or different mutants. The
activity of pmelRcon was determined by b-galactosidase expression and the results are shown in
Figure 2: note that in all of the in vivo assays, the
s70 mutants were expressed in trans to chromosomal rpoD. The results show that pmelRcon is completely dependent upon wild-type CRP for activity
in the presence of wild-type s70 and the different
s70 mutants. The s70 mutants reduce activation by
wild-type CRP to between 54 % and 75 % of that
observed in the presence of wild-type s70. CRP
HL159 KE101, which is defective in both AR1 and
AR2, is de®cient in activation of pmelRcon in the
presence of all the s70 derivatives. With CRP
HL159 KE101 KN52, which activates primarily via
AR3, the presence of either s70 RC596 or RS596
results in similar levels of activation compared
with wild-type s70. However, s70 RG596, s70
Table 1. Derivatives isolated by screening random
mutant libraries of rpoD for candidates that increase
activation by CRP HL159 KE101 KN52 EK58 at
pmelRcon
Amino acid
substitution
Codon substitution
Number of isolates
A. Candidates isolated from a rpoD library in which codons 529 to
613 were randomly mutated
596Arg ! Gly
CGC!GGC
3 from 3 separate
PCR libraries
596Arg ! Ser
CGC!AGC
2 from 2 separate
PCR libraries
596Arg ! Cys
CGC!TGC
1
B. Candidates isolated
degenerate
596Arg ! Gly
596Arg ! Ser
596Arg ! Cys
596Arg ! Glu
596Arg ! Asp
from a rpoD library in which codon 596 was
CGC!GGC
CGC!TCC
CGC!TGC
CGC!GAA
CGC!GAG
CGC!GAC
1
1
1
1
1
2
314
70 Suppressor Mutants
Figure 2. Substitutions at position 596 of s70 increase activation
by CRP HL159 KE101 KN52 EK58
at pmelRcon. The Figure illustrates
b-galactosidase activities (Miller
units) in M182crp cells containing
pRW50 carrying a pmelRcon::lacZ
fusion,
pLG339CRP
encoding
different CRP derivatives and
pVRs encoding different s70
derivatives (see Materials and
Methods).
The
b-galactosidase
activities indicate promoter activity
and the different s70 mutants are
indicated in the legend key. The
bars represent the average of three
independent assays and the error
bars show one standard deviation
either side of the mean.
RE596 or s70 RD596 result in up to a threefold
defect in activation. CRP HL159 KE101 KN52 EK58
is defective in activation of pmelRcon in the presence of wild-type s70. It is striking that all of the
s70 mutants increase activation by CRP HL159
KE101 KN52 EK58 from between 3.6-fold for s70
RC596 to eightfold for the charge reversal derivatives, s70 RE596 and s70 RD596. These results
show that substitutions at R596 of s70 increase activation by CRP HL159 KE101 KN52 EK58 at pmelRcon, and that the charge reversal substitutions give
the largest effects. Thus, in the following experiments we focussed on the effects of s70 RE596 and
s70 RD596 on activation by CRP.
We investigated whether s70 RE596 or s70
RD596 increased activation by a CRP mutant that
contains the positive control substitution, EG58.
This substitution removes the functional side-chain
at position 58 without substituting a positively
charged side-chain. The results in Figure 3 show
that, whilst the charge reversal sigma mutants s70
RE596 and s70 RD596 increase activation at pmelRcon by CRP HL159 KE101 KN52 EK58 compared
with wild-type s70, they decrease activation by
CRP HL159 KE101 KN52 EG58. This demonstrates
that the charge reversal substitutions at position
596 of s70 speci®cally increase activation by CRP
HL159 KE101 KN52 containing K58, but not G58.
To corroborate the results of the in vivo experiments, and to rule out any possibility that the s70
and CRP mutants cause indirect physiological
effects that interfere with the CRP-dependent promoter activities, we performed in vitro single
round transcription assays. To do this, N-terminally His-tagged wild-type s70, s70 RE596 and s70
RD596, and also wild-type CRP and different CRP
mutants were puri®ed using af®nity chromatography (see Materials and Methods). Assays were
then performed using pure RNAP core enzyme
reconstituted with a tenfold excess of s70, s70
RE596 or s70 RD596. The assays consisted of super-
coiled template DNA, containing the pmelRcon promoter cloned upstream of a l oop terminator in the
plasmid, pSR, and either wild-type or mutant CRP.
Figure 4 shows typical transcripts from pmelRcon
in the presence of different CRP and RNAP derivatives, and also transcripts from the reference promoter, RNAI. The CRP-dependent transcripts were
quanti®ed and normalised with respect to transcripts from RNAI. The results show that transcription from pmelRcon by RNAP containing either
wild-type s70, s70 RE596 or s70 RD596 is dependent on CRP. Similar to the in vivo results, activation by wild-type CRP with RNAP containing
either s70 RE596 or s70 RD596 is slightly less than
with wild-type RNAP. Also, CRP HL159 KE101 is
completely defective in activation of all the RNAP
derivatives. CRP HL159 KE101 KN52 activates
transcription in the presence of wild-type RNAP,
but is completely defective in activation of RNAP
containing s70 RE596 or s70 RD596. In contrast,
CRP HL159 KE101 KN52 EK58 is defective in activation of wild-type RNAP, but gives increased
activation with RNAP containing s70 RE596 or s70
RD596. CRP HL159 KE101 KN52 EG58, however,
is defective in activation of all the RNAP derivatives. These in vitro experiments con®rm the in vivo
data and demonstrate that the RE596 or RD596
substitutions in s70 increase activation at pmelRcon
by CRP HL159 KE101 KN52 EK58, but not with
any of the other CRP mutants.
Alanine substitutions at positions 593, 596,
597 and 599 of s 70
The above results show that substitution of R596
of s70 increases activation by CRP HL159 KE101
KN52 EK58. A simple explanation is that substitution of R596 relieves a charge clash created by
the EK58 substitution in AR3. This suggests that
K58 is in close proximity to R596 in s70. Previous
work demonstrated that the positively charged
70 Suppressor Mutants
315
Figure 3. Charge reversal substitutions at position 596 of s70 do
not increase activation by CRP
HL159 KE101 KN52 EG58 at
pmelRcon. The Figure illustrates
b-galactosidase activities (Miller
units) in M182crp cells carrying
pRW50 carrying a pmelRcon::lacZ
fusion,
pLG339CRP
encoding
different CRP derivatives and
pVRs encoding different s70
derivatives (see Materials and
Methods).
The
b-galactosidase
activities indicate promoter activity
and the different s70 mutants are
indicated in the legend key. The
bars represent the average of three
independent assays and the error
bars show one standard deviation
either side of the mean.
residues K593, K597 and R599 in s70 are necessary
for activation by CRP HL159 KE101 KN52, identifying the positively charged residues between positions 593 and 599 as the target of AR3 (Lonetto
et al., 1998). We have investigated whether any of
these positively charged residues are also in close
proximity to K58. To do this, we used transcription
assays to investigate whether single alanine substitutions at positions K593, R596, K597 and R599 in
s70 increase activation by CRP HL159 KE101 KN52
EK58 at pmelRcon. These experiments were performed with pure RNAP core enzyme reconstituted with a tenfold excess of N-terminal GST
tagged s70, s70 KA593, s70 RA596, s70 KA597 or
s70 RA599 (gift from M. Lonetto). Figure 5 shows
typical transcripts from pmelRcon in the presence
of different CRP and RNAP derivatives. The
results show that activation with wild-type CRP is
similar in the presence of the different RNAP
derivatives. With CRP HL159 KE101 there is little
activation with any RNAP. CRP HL159 KE101
KN52, which strongly activates wild-type RNAP
via AR3, gives slightly less activation with RNAP
containing s70 RA596, and gives a twofold defect
in activation with RNAP containing either s70
KA593, s70 KA597 or s70 RA599, con®rming the
result of Lonetto et al. (1998). The weak activation
by CRP HL159 KE101 KN52 EK58 of RNAP is
increased with RNAP containing either s70 RA596
or s70 RA599. In contrast, the weak activation by
CRP HL159 KE101 KN52 EG58 of wild-type RNAP
is not increased with any of the mutant RNAP
holoenzymes. This suggests that the s70 substitutions RA596 and RA599 speci®cally relieve a
clash introduced by the EK58 substitution in AR3.
However, the effects observed with s70 RA596 and
s70 RA599 are much weaker than with the charge
reversal mutants, s70 RE596 and s70 RD596.
Activity of the s 70 suppressor mutants in the
context of wild-type CRP
The EK58 and EG58 substitutions in AR3 also
decrease activation by wild-type CRP, in which
AR1 and AR2 are fully functional and AR3 is not
improved (V.R. unpublished data). Thus, here we
investigated whether the strongest s70 suppressor
mutants, s70 RE596 and s70 RD596, also increase
activation by CRP EK58 at a Class II promoter.
Since both EK58 and EG58 in wild-type CRP result
in greater defects in activation at the CRP-dependent Class II promoter CC(ÿ41.5) compared with
pmelRcon (Rhodius & Busby, 2000; V.R. unpublished data), we used transcription assays to
measure activation by CRP EK58 using the supercoiled template, pSR/CC(ÿ41.5). In addition,
assays were performed with wild-type CRP, CRP
HL159, which is defective in AR1, CRP KE101,
which is defective in AR2, CRP EG58 and CRP
KA52, in which the inhibitory K52 side-chain is
removed. The results in Figure 6 show that transcription from CC(ÿ41.5) by RNAP containing
either wild-type s70, s70 RE596 or s70 RD596 is
dependent on CRP. Activation by wild-type CRP
with wild-type RNAP is reduced twofold with
RNAP containing either s70 RE596 or s70 RD596,
whilst with the positive control mutants, CRP
HL159 and CRP KE101, very little activation occurs
with any of the RNAP derivatives. CRP EK58
results in 10 % activation with wild-type RNAP,
but strikingly, this defect is almost completely
restored with RNAP containing either s70 RE596
or s70 RD596. In contrast, CRP EG58 results in
50 % activation with wild-type RNAP, and the
RE596 or RD596 substitutions in s70 have little
effect. The KA52 substitution in CRP results in a
®vefold increase in transcription with wild-type
RNAP (compared with wild-type CRP). It is inter-
316
70 Suppressor Mutants
Figure 4. In vitro transcription
assays showing the effect of the s70
charge reversal substitutions at
position 596 on CRP-dependent
activation from pmelRcon. The
Figure shows a typical gel of radiolabelled transcripts from single
round in vitro transcription assays.
The experiment was performed
with supercoiled template containing pmelRcon cloned in pSR, in the
presence of different CRP derivatives and RNAP reconstituted with
different N-terminal His6 tagged
s70 derivatives (see Materials and
Methods). CRP-dependent transcripts from pmelRcon and control
transcripts from the RNAI promoter are indicated, and different
His6-s70 and CRP derivatives are
labelled below each lane. The histogram illustrates the amount of
CRP-dependent transcript produced from pmelRcon. The data are
presented as ``percentage of activation by wild-type CRP in the
presence of wild-type His6-s70,
normalised to RNAI'', which is:
100 ((CRP-dependent transcript in the presence of CRP and His6-s70 derivative/RNAI) ÿ (CRP-independent transcript with His6-s70 derivative/RNAI))/((CRP-dependent transcript in the presence of wild-type CRP and His6-s70/
RNAI) ÿ (CRP-independent transcript with His6-s70/RNAI)). The bars in the histogram represent the average of
three independent assays and the error bars show one standard deviation either side of the mean.
esting that this increase is not observed with
RNAP containing either s70 RE596 or s70 RD596.
Our work has shown that substitutions in AR3
of CRP do not affect activation from Class I CRPdependent promoters (Rhodius & Busby, 2000).
Thus, as controls, transcription assays were performed with the different CRP and RNAP mutants
at the Class I CRP-dependent promoter, CC(ÿ61.5),
using the supercoiled template, pSR/CC(ÿ61.5).
The results in Figure 7 show that transcription
from CC(ÿ61.5) is dependent on wild-type CRP. In
addition, CRP-dependent transcription requires
AR1, but not AR2 or AR3 of CRP. The RE596 and
RD596 substitutions in s70 hardly affect activation
by any of the CRP mutants.
Conclusions
s 70 determinants that contact AR3
Lonetto et al. (1998) demonstrated that residues
K593, K597 and R599 in s70 are required for AR3dependent activation by CRP. These positively
charged residues of s70 provide a complementary
target for the negatively charged activatory determinant in AR3 of CRP composed of residues D53,
E54, E55 and E58. Within this determinant, residue
E58 is the single most important side-chain
required for AR3-dependent activation (Rhodius &
Busby, 2000). We predicted that the charge reversal
substitution in AR3, EK58, results in a decrease in
activation due to a charge clash with one or more
targets in s70. We have identi®ed s70 mutants that
increase activation by CRP derivatives containing
the EK58 substitution. Our results establish that
substitutions of either R596 or R599 increase activation by CRP HL159 KE101 KN52 EK58. Our
results show that the largest and clearest effects are
with charge reversal substitutions at R596 in s70.
The speci®city of the effects of these charge reversal substitutions provides good evidence for a
direct interaction between residue 58 in AR3 of
CRP and residue 596 in the C-terminal region of
s70 at Class II CRP-dependent promoters. First,
these substitutions in s70 enhance activation by
CRP HL159 KE101 KN52 EK58 and CRP EK58, but
do not enhance activation by CRP HL159 KE101
KN52 EG58 and CRP EG58. Second, they do not
enhance activation by CRP mutants defective in
either AR1 and/or AR2. Third, they decrease activation by wild-type CRP and CRP HL159 KE101
KN52. Finally, the charge reversal substitutions of
R596 in s70 do not alter activation by any CRP
mutants at the Class I CRP-dependent promoter,
CC(ÿ61.5). We suggest that replacement of R596 in
s70 with Glu or Asp residues re-educates a contact
with K58 in AR3 by charge-charge or direct hydrogen bond interactions. In the case of CRP containing the EG58 substitution, both s70 RE596 and s70
RD596 are unable to make a functional interaction
with G58, thus no increase in activation is
observed. In the cases of CRP derivatives contain-
70 Suppressor Mutants
317
Figure 5. In vitro transcriptions showing the effect of different s70 alanine mutants on CRP-dependent activation
from pmelRcon. The Figure shows a typical gel of radiolabelled transcripts from single round in vitro transcription
assays. The experiment was performed with supercoiled template containing pmelRcon in the presence of different
CRP derivatives and RNAP reconstituted with different N-terminal GST tagged s70 mutants (see Materials and
Methods). CRP-dependent transcripts from pmelRcon and control transcripts from the RNAI promoter are indicated,
and different GST-s70 and CRP derivatives are labelled below each lane. The histogram illustrates the amount of
CRP-dependent transcript produced from pmelRcon. For each CRP derivative, the data are presented as ``a ratio of
activation in the presence of mutant GST-s70 divided by activation in the presence of wild-type GST-s70, normalised
to RNAI'', which is: ((pmelRcon in the presence of CRP derivative/RNAI) ÿ (pmelRcon in the absence of CRP/RNAI))
for each GST-s70 mutant, divided by the equivalent expression for wild-type GST-s70. The bars represent the average
of three independent assays and the error bars show one standard deviation either side of the mean.
ing the native residue E58, we propose that a
charge-charge clash occurs with E596 or D596 in
s70, resulting in a decrease in activation. It is interesting that the screen of the s70 random mutant
library for candidates that increase activation by
CRP HL159 KE101 KN52 EK58 only identi®ed
mutants that contained substitutions at position
596. However, an alanine substitution at R599 in
s70 also increased activation by CRP HL159 KE101
KN52 EK58. We conclude that both R596 and R599
in s70 are close to the side-chain of residue 58 in
AR3 of CRP at Class II promoters. It is likely that
substitutions of R599 were too weak to be detected
in our in vivo screen.
Residue K52 of AR3 is inhibitory to maximal
levels of activation, such that substitution of the
lysine for other residues results in an increase in
Class II CRP-dependent activation (Rhodius &
Busby, 2000). The mechanism of inhibition by K52
is not understood, but one possibility is that K52 of
CRP clashes with the positively charged target on
s70. Thus, substitution of K52 for an alanine
relieves this clash resulting in an increase in acti-
vation. Our results show that the KA52 substitution has no effect on CRP-dependent activation
of RNAP containing either s70 RE596 or s70
RD596. Thus, we conclude that K52 in AR3 clashes
with R596 in s70 at Class II promoters.
A model for interactions between AR3 and s 70
Based on our results, it is possible to build a
speculative model of the interactions between the
activatory and inhibitory determinants in AR3 and
the target in s70. We propose that E58 is in close
proximity to R596 and R599 in s70. We also propose that K52 is in close proximity to R596 in s70,
thus generating a steric and/or a charge-charge
clash. This leaves the negatively charged cluster of
residues, D53, E54 and E55, located at the apex of
the b-turn of AR3. We suggest that they interact
with the remaining determinants on s70 required
for AR3-dependent activation identi®ed by Lonetto
et al. (1998), residues K593, R597 and R599. This is
illustrated schematically in Figure 8.
318
70 Suppressor Mutants
Figure 6. In vitro transcriptions showing the effect of the s70 charge reversal substitutions at position 596 on CRPdependent activation from CC(ÿ41.5). The Figure shows a typical gel of radiolabelled transcripts from single round
in vitro transcription assays. The experiment was performed with supercoiled template containing CC(ÿ41.5) cloned
in pSR, in the presence of different CRP derivatives and RNAP reconstituted with different N-terminal His6 tagged
s70 mutants. CRP-dependent transcripts from CC(ÿ41.5) and control transcripts from the RNAI promoter are indicated, and different His6-s70 and CRP derivatives are labelled below each lane. The histogram illustrates the amount
of CRP-dependent transcript produced from CC(ÿ41.5). The data are presented as in Figure 4.
The structure of the C-terminal region of s70 has
not yet been solved. Based on homology modelling,
Lonetto et al. (1998) proposed two possible structures for s70 region 4 based on the helix-turn-helix
motifs of NarL and 434 Cro. The principal difference between these two structures is the length of
the second helix of the helix-turn-helix motif (recognition helix). In the NarL-based model, the target
residues for AR3 are located at the C-terminal end
of an extended recognition helix. In contrast, in the
434 Cro-based model the recognition helix is shorter, such that the target residues for AR3 are located
on a large surface-exposed loop following the helix.
Based on our evidence that AR3 interacts with the
positively charged residues in s70, we tested the
models to see which would provide the best alignment between s70 and AR3. We assumed that, at
Class II CRP-dependent promoters, region 4 of s70
is docked in the ÿ35 hexamer of the promoter in a
similar manner to that of activator-independent
promoters (Bown et al., 2000). Thus, we found that
the NarL model best aligned residues 593, 596, 597
and 599 in s70 with determinants in AR3 (Figure 9).
This suggests that the extended recognition helix
model provided by NarL may serve as a more accurate template for region 4 of s70.
Other Class II activators also interact with
R596 of s 70
Substitutions at position 596 of s70 have both
positive and negative effects on activation by other
transcription factors that bind to DNA sites that
overlap the ÿ35 hexamer of target promoters. For
example, at the araBAD promoter, which is
co-regulated by AraC and CRP, substitutions
changing R596 of s70 to His, Cys, Ser or Ala enable
AraC to activate transcription in the absence of
CRP (Hu & Gross, 1985; Lonetto et al., 1998). Likewise, with the lcI protein, s70 RH596 increases
activation at PRM by a lcI positive control mutant
carrying the substitution, DN38 (Li et al., 1994). In
contrast, His, Cys and Ser substitutions at position
596 of s70 reduce MalT-dependent activation of
PmalK-lamB (Hu & Gross, 1985), and Lonetto et al.
(1998) demonstrated that FNR requires R596 for
activation of PdmsA, and also PnarG in the absence
of the co-regulator, NarL. In addition, at the
Bacillus subtilis sH-dependent promoter, spoIIA, the
Spo0A transcription factor requires the equivalent
residue of sH, R205, to activate transcription
(Buckner & Moran, 1998). Thus R596 in s70 plays
an important role in transcription activation at a
variety of activator-dependent promoters. It is
70 Suppressor Mutants
319
Figure 7. In vitro transcriptions showing the effect of the s70 charge reversal substitutions at position 596 on CRPdependent activation from CC(ÿ61.5). The Figure shows a typical gel of radiolabelled transcripts from single round
in vitro transcription assays. The experiment was performed with supercoiled template containing CC(ÿ61.5) cloned
in pSR, in the presence of different CRP derivatives and RNAP reconstituted with different N-terminal His6 tagged
s70 mutants. CRP-dependent transcripts from CC(ÿ61.5) and control transcripts from the RNAI promoter are indicated, and different His6-s70 and CRP derivatives are labelled below each lane. The histogram illustrates the amount
of CRP-dependent transcript produced from CC(ÿ61.5). The data are presented as in Figure 4.
likely that R596 is located within a larger surfaceexposed target site within the C-terminal region of
s70, and thus is able to make direct protein-protein
interactions with many activators.
Materials and Methods
Strains, plasmids and recombinant
DNA methodology
The bacterial strains, plasmids and promoter fragments used in this study are listed in Table 2. The E. coli
strain BLR(DE3) pLysS is a RecAÿ strain carrying the
inducible T7 RNA polymerase gene on the chromosome
and expresses low levels of T7 lysozyme from the plasmid, pLysS. BLR(DE3) pLysS was used to express Histagged s70 derivatives from the plasmid, pET-21s, and
the T7 lysozyme ensured little or no expression of Histagged s70 before induction. CRP derivatives are listed
in Table 3. Standard methods for isolation and manipulation of DNA fragments were used throughout
(Sambrook et al., 1989). Synthetic oligonucleotides used
either for sequencing, for PCR or for constructions were
purchased from Alta Bioscience at the University of
Birmingham. Bacterial strains carrying different plasmids were propagated in LB or on MacConkey lactose
plates containing 35 mg/ml tetracycline, 25 mg/ml
kanamycin or 80 mg/ml ampicillin, as appropriate.
The strain BLR(DE3) pLysS was maintained on LB
plates containing 12.5 mg/ml tetracycline and 34 mg/ml
chloramphenicol.
For the in vitro transcription assays, the vector pSR
was used carrying the CC(ÿ41.5), CC(ÿ61.5) or pmelRcon
promoters on short EcoRI-HindIII fragments in which the
HindIII site is located at position ‡36 downstream of the
transcription start (Rhodius & Busby, 2000).
The plasmid, pVRs, is a pBR322-based rpoD
expression vector derived from the plasmid pKBs (Barne
et al., 1997). In pKBs, rpoD is expressed by the CRPregulated promoter, galP1. This promoter was replaced
by digesting pKBs with EcoRI and XcmI to remove the
galP1 promoter sequence and replacing it with an EcoRIXcmI fragment containing the constitutive promoter,
galP1-27 (Busby et al., 1987). Similar to galP1, the
galP1-27 promoter contains an extended `` ÿ 10`` motif,
but instead has all the sequences upstream of ÿ27 from
the transcription start point removed, thereby creating a
CRP-independent, but constitutive s70 promoter. In
addition, to create pVRs, the galK gene was partially
removed from the religated vector by digesting with
NarI and AccI, treating with Klenow enzyme and
religating the blunt vector ends.
The plasmid pET-21s is a T7lac expression plasmid
for N-terminal His-tagged s70 derivatives, in which the
rpoD gene was originally obtained from the T7
expression vector, pGEMD (Igarashi & Ishihama, 1991).
pGEMD was reconstructed by J. Bown to pGEMHisD,
which expresses N-terminally His-tagged s70 that contains a thrombin cleavage site between the hexa-His tag
and Met1 of s70. The His tag and thrombin cleavage site
sequence (amino acid sequence MGSSH6SSGLVPRGSH)
was obtained from the vector, pET-15b (Novagen, UK),
by PCR ampli®cation using a primer that anneals
upstream of the XbaI site and a second mutagenic
320
70 Suppressor Mutants
Table 2. Bacterial strains, plasmids and promoters used in this work
Brief description
A. Bacterial strains
M182crp
DH5a
BLR(DE3) pLysS
B. Plasmids
RK2 replication origin encoding TetR
pRW50
ColE1 replication origin encoding AmpR
pSR
pVRs and derivatives
pDCRP and derivatives
f1 replication origin encoding AmpR
pET-21s and derivatives
pSC101 replication origin encoding KanR
pLG339CRP and derivatives
Origin
E. coli K12 lac crp
E. coli DH5a
ÿ
E. coli Fÿ ompT hsdSB(rÿ
B mB ) gal dcm (srlrecA)306::Tn10(DE3) pLysS
Busby et al. (1983)
Hanahan (1983)
Novagen, UK
Broad host range low copy lac expression
vector for cloning EcoRI-HindIII promoter
fragments
Lodge et al. (1992)
pBR322 derivative containing a loop
transcription terminator downstream of
promoter sequences cloned on EcoRI-HindIII
fragments
pBR322 derivative encoding rpoD and mutant
derivatives
pBR322 derivative encoding crp and mutant
derivatives (see Table 3)
Kolb et al. (1995)
This work
West et al. (1993)
pET-21a(‡) based over expression vector
encoding rpoD with an N-terminal hexa-His
tag under the inducible control of the T7lac
promoter
This work
Low copy number plasmid (previously
referred to as pDW300) encoding crp and
mutant derivatives (see Table 3)
West et al. (1993)
C. Promoters
(cloned on EcoRI-HindIII fragments in pRW50 and pSR)
pmelRcon
Derivative of the E. coli melR promoter with
point mutations at ÿ45 and ÿ49 that improve
the CRP-binding site
CC(ÿ41.5)a
Class II derivative of the melR promoter with a
consensus CRP-binding site centred at ÿ41.5
a
CC(ÿ61.5)
Class I derivative of the melR promoter with a
consensus CRP-binding site centred at ÿ61.5
West et al. (1993)
Gaston et al. (1990)
Gaston et al. (1990)
a
In previous papers (e.g. Gaston et al., 1990), CC(ÿ41.5) was referred to as CCpmelR. The nomenclature has been changed to harmonise with Zhou et al. (1994a,b) and with more recent publications.
primer, 50 -GTTTCAGCTGTGACTGCGGGTTTTGCTCCATATGGCTGCCGCGCGGCACCAGGCCGC-30 . The 30
segment of the mutagenic primer (underlined) is complementary to the non-coding strand equivalent to the
thrombin cleavage site in pET-15b, and the 50 segment is
complementary to the non-coding strand equivalent to
codons 1 to 11 of rpoD in pGEMD and encompasses a
PvuII restriction site (double underlined). The PCR
products were digested with XbaI and PvuII, and ligated
into pGEMD XbaI-PvuII vector, to make pGEMHisD.
Segments of rpoD encoding residues 529 to 613 and the
substitutions RE596 or RD596 were transferred from
pVRs on XhoI-HindIII fragments into pGEMHisD XhoIHindIII vector to make pGEMHisD RE596 and pGEMHisD RD596. The T7 promoter of pGEMHisD allows
residual expression of rpoD when transformed into the
T7 expression strain BLR(DE3) pLysS and grown in LB
under non-induced conditions. To combat this, XbaIHindIII fragments encoding rpoD were transferred from
pGEMHisD, pGEMHisD RE596 and pGEMHisD RD596
and cloned into pET-21a(‡) XbaI-HindIII vector, such
that they were placed under the control of the inducible
T7lac promoter. This gave the ®nal constructs pET-21s,
pET-21s RE596 and pET-21s RD596, which were used
to overexpress the different s70 derivatives in BLR(DE3)
pLysS.
Random mutagenesis of the rpoD coding region
Error-prone PCR (Zhou et al., 1991) was used to prepare a library of random mutations in the 30 region of
the rpoD gene encoding residues 530 to 613. The rpoD
coding region of pVRs was ampli®ed by PCR using Taq
polymerase and primers which ¯anked the XhoI restriction site located at codon 529 in rpoD and the HindIII
restriction site just downstream of the rpoD gene. The
PCR products were digested with XhoI and HindIII and
ligated into pVRs XhoI-HindIII vector to generate a
library of derivatives carrying random mutations in
the 30 region of rpoD. The library was transformed
into M182crp cells carrying pRW50/pmelRcon and
pLG339CRP encoding CRP HL159 KE101 KN52. The
transformants were plated on MacConkey agar indicator
plates containing 10 g/l lactose, 35 mg/ml tetracycline,
25 mg/ml kanamycin and 80 mg/ml ampicillin.
A second library of rpoD derivatives, in which codon
596 was completely randomised, was constructed in the
vector pVRs using megaprimer PCR (Perrin & Gilliland,
1990). The ®rst round of PCR used a primer that
321
70 Suppressor Mutants
Table 3. crp derivatives used in this work
Brief description
Figure 8. Schematic model of interactions between
AR3 and s70. The Figure illustrates proposed interactions between residues in AR3 of CRP and target residues in the C-terminal region of s70 in the
transcriptional complex at a Class II CRP-dependent
promoter. Residues in AR3 are displayed on a surface
exposed b-turn, with the large arrowhead indicating the
direction of the peptide chain towards the C terminus.
Based on the crystallographic structure of the CAPDNA complex (Schultz et al., 1991; Parkinson et al.,
1996), residues D53, E54 and E55 form a negatively
charged cluster at the apex of the b-turn, and the sidechains of K52 and E58 are adjacent to each other. Residues 593 to 597 of s70 are located on an a-helix, and
R599 on a surface exposed loop, based on homology
modelling of the helix-turn-helix motif of s70 with NarL
(Baikalov et al., 1996; Lonetto et al., 1998). The a-helix is
viewed from the carboxyl end and the loop is illustrated
with the large arrowhead indicating the direction of the
peptide chain towards the C terminus. The side-chains
of K593, K597 and R599 form a cluster, separate from
R596. Productive interactions between residues are indicated by double headed arrows, non-productive interactions (line with perpendicular ends) and weak
interactions (dotted line).
annealed downstream of the HindIII site located just
downstream of rpoD in pVRs, and a second mutagenic
primer that annealed to the coding strand of rpoD such
that the primer sequence for codon 596 was completely
degenerate. This created a megaprimer that was used in
a second round of PCR together with a primer that
annealed upstream of the XhoI site located at codon 529
of rpoD in pVRs. The ®nal PCR product was digested
with XhoI and HindIII and ligated into pVRs XhoIHindIII vector to generate a library of rpoD derivatives
in which codon 596 was completely randomised. The
library was transformed into the tester strain described
above.
Measurement of promoter activity in vivo
Expression of different promoter::lac fusions in vivo in
the presence of different CRP and s70 derivatives was
determined using the b-galactosidase assay method of
Miller (1972). M182crp cells containing pRW50/
pmelRcon carrying the pmelRcon::lacZ fusion, pLG339CRP
carrying the appropriate crp derivative and pVRs carrying the appropriate rpoD derivative were grown aerobically at 37 C to mid-log phase in LB with the
appropriate antibiotics. b-Galactosidase activities were
determined as described previously (Bell et al., 1990;
Lodge et al., 1992) and the values (in Miller units) taken
Origin
CRP derivatives (cloned on EcoRI-HindIII fragments in pDCRP
and BamHI-SalI fragments in pLG339CRP)
CRP
Wild-type CRP
CRP HL159
CRP with defective Rhodius et al. (1997)
AR1
CRP KE101
CRP with defective Rhodius et al. (1997)
AR2
CRP HL159 KE101 CRP with defective Rhodius et al. (1997)
AR1 and AR2
CRP HL159 KE101 CRP with defective Lonetto et al. (1998)
KN52
AR1 and AR2, and
an improved AR3
CRP HL159 KE101 Derivative of CRP Rhodius & Busby
KN52 EK58
HL159 KE101 KN52 (2000)
with a disrupted
AR3
CRP HL159 KE101 Derivative of CRP Rhodius & Busby
KN52 EG58
HL159 KE101 KN52 (2000)
with a disrupted
AR3
CRP KA52
CRP with an
Rhodius & Busby
improved AR3
(2000)
CRP EK58
CRP with defective Rhodius & Busby
AR3
(2000)
CRP EG58
CRP with defective Rhodius & Busby
AR3
(2000)
to be proportional to promoter activity. We noted that
cultures of M182crp carrying either pVRs RE596 or
pVRs RD596 in the presence of pDCRP grew extremely
slowly and sometimes lysed, presumably due to the toxicity of the mutant rpoD gene products.
Purification of CRP and s 70 derivatives
CRP proteins were puri®ed from cultures of
M182crp cells transformed with pDCRP encoding the
appropriate crp derivatives following the method of
Ghosaini et al. (1988). The N-terminal GST-tagged wildtype s70 and derivatives KA593, RA596, KA597 and
RA599 were generously donated by M. Lonetto (Lonetto
et al., 1998).
The N-terminal His-tagged wild-type s70 and derivatives RE596 and RD596 were puri®ed from cultures of
the T7 over expression strain, BLR(DE3) pLysS, carrying
the expression vector, pET-21s that encodes T7lac::rpoD,
described above. The s70 proteins form inclusion bodies
in the overproducing strain and were puri®ed under
denaturing conditions using a Nickel af®nity matrix
(Ni-NTA Agarose purchased from QIAGEN, UK). Cultures were grown aerobically at 37 C in LB containing
12.5 mg/ml tetracycline, 34 mg/ml chloramphenicol and
200 mg/ml ampicillin to A650 ˆ 0.5 ÿ 0.6, before inducing
with 1 mM IPTG. After incubation for a further three
hours, the cells were harvested and stored at ÿ70 C.
The frozen cells were resuspended in 10 mM Tris-HCl
(pH 7.6 at 4 C) buffer containing 1 mM EDTA and lysed
by sonication before centrifuging at 10,000 g for ten minutes at 4 C. The pellet was washed by resuspending in
50 mM Tris-HCl (pH 7.6 at 4 C), 500 mM NaCl and
0.5 % (v/v) Triton and centrifuging again at 10,000 g for
ten minutes at 4 C. The pellet containing the s70 subunit
was solubilized with GDHCl buffer (6 M guanidine-HCl
in 10 mM Tris-HCl (pH 7.6 at 4 C) and 500 mM NaCl)
‡ 10 mM imidazole and applied to a pre-equilibrated
322
70 Suppressor Mutants
Figure 9. Model of CRP and s70 C-terminal region showing alignment of contact patches. The model illustrates the
proximity of CRP AR3 and the target residues of s70 when CRP is bound to a target site centred at position ÿ41.5 at
Class II promoters. The CRP dimer-DNA model (yellow) is based on the crystallographic structure of the CRP-DNA
complex (Parkinson et al., 1996) and residues 551 to 613 of s70 (grey) are based on the helix-turn-helix structure of
NarL (Lonetto et al., 1998). The DNA binding helix-turn-helix motif of s70 is ``approximately docked'' in the major
groove of the ÿ35 region of the promoter. Residues K52, D53, E54, E55 and E58 of AR3 (red), residues K593, K597
and R599 of s70 (dark blue), and residue R596 of s70 (light blue).
Ni-NTA Agarose column at 4 C. The column was
washed with ten column volumes of GDHCl buffer
‡ 10 mM imidazole and the His-tagged s70 subunits
eluted in GDHCl buffer with an increasing stepwise
gradient of 20-100 mM imidazole. Fractions containing
s70 were dialysed against 10 mM Tris-HCl (pH 7.6 at
4 C), 10 mM MgCl2, 0.1 mM EDTA, 50 % (w/v) glycerol, 0.1 M KCl, 0.1 mM DTT and then stored at ÿ20 C.
s70 puri®ed by this method was 99 % pure as judged by
SDS PAGE.
In vitro transcription assays
The single round in vitro transcription assays were
performed exactly as described in Lonetto et al. (1998).
The DNA templates, pSR/pmelRcon, pSR/CC(ÿ41.5) and
pSR/CC(ÿ61.5), were puri®ed by CsCl gradient centrifugation (Sambrook et al., 1989). Binding reactions contained 5 nM template DNA, 25 nM core RNAP (supplied by Epicentre Technologies, UK) saturated with a
tenfold excess of wild-type or mutant s70, and wild-type
or mutant CRP (50 nM with templates containing the
CC(ÿ41.5) or CC(ÿ61.5) promoters, or 250 nM with
pmelRcon). Radiolabelled transcripts were electrophoresed on denaturing 6 % (w/v) polyacrylamide gels and
quanti®ed on a Molecular Dynamics phosphorimager
using the software, ImageQuant, v3.3. The CRP-dependent transcripts were normalised against the control
RNAI transcript and also against an end-labelled oligonucleotide included in the reaction loading buffer to
compensate for variations in gel loading.
Acknowledgements
We thank Mike Lonetto and Carol Gross for donating
the GST tagged s70 proteins, Jon Bown for the gift of the
plasmid, pGEMHisD, and Nigel Savery for puri®cation
of some of the CRP derivatives. We are grateful to Jon
Bown, Richard Ebright, Carol Gross, Ann Hochschild,
and Nigel Savery for helpful discussions. This work was
generously supported by grants 044764 and 055993 from
the Wellcome Trust to S.B.
References
Baikalov, I., Schroder, I., KaczorGrzeskowiak, M.,
Grzeskowiak, K., Gunsalus, R. P. & Dickerson, R. E.
(1996). Structure of the Escherichia coli response
regulator NarL. Biochemistry, 35, 11053-11061.
Barne, K., Bown, J., Busby, S. & Minchin, S. (1997).
Region 2.5 of the Escherichia coli RNA polymerase
s70 subunit is responsible for the recognition of the
70 Suppressor Mutants
``extended ÿ10`` motif at promoters. EMBO J. 16,
4034-4040.
Bell, A., Gaston, K., Williams, R., Chapman, K., Kolb,
A., Buc, H., Minchin, S., Williams, J. & Busby, S.
(1990). Mutations that alter the ability of the Escherichia coli cyclic AMP receptor protein to activate
transcription. Nucl. Acids Res. 18, 7243-7250.
Bown, J. A., Kolb, A., Meares, C. F., Ishihama, A.,
Minchin, S. D. & Busby, S. J. W. (2000). Positioning
region 4 of the Escherichia coli RNA polymerase s70
subunit by a transcription activator. J. Bacteriol. In
the press.
Buckner, C. M. & Moran, C. P., Jr. (1998). A region in
Bacillus subtilis sH required for Spo0A-dependent
promoter activity. J. Bacteriol. 180, 4987-4990.
Busby, S. & Ebright, R. (1997). Transcription activation
at Class II CAP-dependent promoters. Mol. Microbiol. 23, 853-859.
Busby, S. & Ebright, R. (1999). Transcription activation
by catabolite activator protein (CAP). J. Mol. Biol.
293, 199-213.
Busby, S., Kotlarz, D. & Buc, H. (1983). Deletion mutagenesis of the Escherichia coli galactose operon
promoter region. J. Mol. Biol. 154, 211-227.
Busby, S., Spassky, A. & Chan, B. (1987). RNA polymerase makes important contacts upstream from basepair ÿ49 at the Escherichia coli galactose operon P1
promoter. Gene, 53, 145-152.
Chan, B. & Busby, S. (1989). Recognition of nucleotide
sequences at the Escherichia coli galactose operon P1
promoter by RNA polymerase. Gene, 84, 227-236.
Gaston, K., Bell, A., Kolb, A., Buc, H. & Busby, S.
(1990). Stringent spacing requirements for transcription activation by CRP. Cell, 62, 733-743.
Ghosaini, L. R., Brown, A. M. & Sturtevant, J. M. (1988).
Scanning calorimetric study of the thermal unfolding of catabolite activator protein from Escherichia
coli in the absence and presence of cyclic mononucleotides. Biochemistry, 27, 5257-5261.
Gross, C., Chan, C., Dombroski, A., Gruber, T., Sharp,
M., Tupy, J. & Young, B. (1998). The functional and
regulatory roles of sigma factors. In Transcription.
Cold Spring Harbor Symposia on Quantitative Biology,
LXIII, pp. 141-155, Cold Spring Harbor Laboratory
Press, Cold Spring Harbor, NY.
Hanahan, D. (1983). Studies on the transformation of
Escherichia coli with plasmids. J. Mol. Biol. 166, 557580.
Hu, J. C. & Gross, C. A. (1985). Mutations in the sigma
subunit of E. coli RNA polymerase which affect
positive control of transcription. Mol. Gen. Genet.
199, 7-13.
Igarashi, K. & Ishihama, A. (1991). Bipartite functional
map of the E. coli RNA polymerase a subunit:
involvement of the C-terminal region in transcription activation by cAMP-CRP. Cell, 65, 1015-1022.
Jin, R., Sharif, K. & Krakow, J. (1995). Evidence for contact between the cyclic AMP receptor protein and
the s70 subunit of Escherichia coli RNA polymerase.
J. Biol. Chem. 270, 19213-19216.
Kim, S.-K., Makino, K., Amemura, M., Nakata, A. &
Shinagawa, H. (1995). Mutational analysis of the
role of the ®rst helix of region 4. 2 of the s70 subunit of Escherichia coli RNA polymerase in transcriptional activation by activator protein PhoB. Mol.
Gen. Genet. 248, 1-8.
Kolb, A., Kotlarz, D., Kusano, S. & Ishihama, A. (1995).
Selectivity of the Escherichia coli RNA polymerase
323
Es38 for overlapping promoters and ability to support CRP activation. Nucl. Acids Res. 23, 819-826.
Kuldell, N. & Hochschild, A. (1994). Amino acid substitutions in the 35 recognition motif of s70 that result
in defects in phage l repressor-stimulated transcription. J. Bacteriol. 176, 2991-2998.
Kumar, A., Malloch, R., Fujita, N., Smillie, D., Ishihama,
A. & Hayward, R. (1993). The minus 35-recognition
region of Escherichia coli sigma 70 is inessential for
initiation of transcription at an ``extended minus
10`` promoter. J. Mol. Biol. 232, 406-418.
Landini, P. & Busby, S. J. W. (1999). The Escherichia coli
ada protein can interact with two distinct determinants in the s70 subunit of RNA polymerase
according to promoter architecture: identi®cation of
the target of ada activation at the alkA promoter.
J. Bacteriol. 181, 1524-1529.
Landini, P., Bown, J. A., Volkert, M. R. & Busby, S. J. W.
(1998). Ada protein-RNA polymerase s subunit
interaction and a subunit-promoter DNA interaction are necessary at different steps in transcription initiation at the Escherichia coli ada and aidB
promoters. J. Biol. Chem. 273, 13307-13312.
Li, M., Moyle, H. & Susskind, M. M. (1994). Target of
the transcriptional activation function of phage lcI
protein. Science, 263, 75-77.
Lodge, J., Fear, J., Busby, S., Gunasekaran, P. & Kamini,
N.-R. (1992). Broad host range plasmids carrying
the Escherichia coli lactose and galactose operons.
FEMS Letters, 95, 271-276.
Lonetto, M., Rhodius, V., Lamberg, K., Kiley, P., Busby,
S. & Gross, C. (1998). Identi®cation of a contact site
for different transcription activators in region 4 of
the Escherichia coli RNA polymerase s70 subunit.
J. Mol. Biol. 284, 1353-1365.
Miller, J. (1972). Experiments in Molecular Genetics, Cold
Spring Harbor Laboratory Press, Cold Spring
Harbor, NY.
Parkinson, G., Wilson, C., Gunasekera, A., Ebright, Y.,
Ebright, R. & Berman, H. (1996). Structure of the
Ê resolution. J. Mol. Biol.
CAP-DNA complex at 2. 5 A
243, 595-602.
Perrin, S. & Gilliland, G. (1990). Site-speci®c mutagenesis using asymmetric polymerase chain reaction
and a single mutant primer. Nucl. Acids Res. 18,
7433-7438.
Rhodius, V. & Busby, S. (1998). Positive activation of
gene expression. Curr. Opin. Microbiol. 1, 152-159.
Rhodius, V. & Busby, S. (2000). Transcription activation
by the Escherichia coli cyclic AMP receptor protein:
determinants within activating region 3. J. Mol. Biol.
299, 295-310.
Rhodius, V., West, D., Webster, C., Busby, S. & Savery,
N. (1997). Transcription activation at Class II CRPdependent promoters: the role of different activating regions. Nucl. Acids Res. 25, 326-332.
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989).
Molecular Cloning. A Laboratory Manual, 2nd edn.,
Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY.
Schultz, S., Shields, G. & Steitz, T. (1991). Crystal structure of a CAP-DNA complex: the DNA is bent by
90 degrees. Science, 253, 1001-1007.
West, D., Williams, R., Rhodius, V., Bell, A., Sharma, N.,
Zou, C., Fujita, N., Ishihama, A. & Busby, S. (1993).
Interactions between the Escherichia coli cyclic AMP
receptor protein and RNA polymerase at Class II
promoters. Mol. Microbiol. 10, 789-797.
324
70 Suppressor Mutants
Williams, R., Bell, A., Sims, G. & Busby, S. (1991). The
role of two surface exposed loops in transcription
activation by the Escherichia coli CRP and FNR
proteins. Nucl. Acids Res. 19, 6705-6712.
Williams, R., Rhodius, V., Bell, A., Kolb, A. & Busby, S.
(1996). Orientation of functional activating regions
in the Escherichia coli CRP protein during transcription activation at class II promoters. Nucl. Acids Res.
24, 1112-1118.
Zhou, Y., Zhang, X. & Ebright, R. (1991). Random mutagenesis of gene-sized DNA molecules by use of
PCR with Taq DNA polymerase. Nucl. Acids Res. 19,
6052.
Zhou, Y., Merkel, T. & Ebright, R. (1994a). Characterization of the activating region of Escherichia coli catabolite gene activator protein (CAP) II. Role at Class
I and Class II CAP-dependent promoters. J. Mol.
Biol. 243, 603-610.
Zhou, Y., Pendergrast, P., Bell, A., Williams, R., Busby,
S. & Ebright, R. (1994b). The functional subunit of a
dimeric transcription activator protein depends on
promoter architecture. EMBO J. 13, 4549-4557.
Edited by R. Ebright
(Received 4 January 2000; received in revised form 13 March 2000; accepted 24 March 2000)