Many transcription factors, including the Escherichia coli cyclic AMP receptor protein (CRP), act by making direct contacts with RNA polymerase. At Class II CRP‐dependent promoters, CRP activates transcription by making two such contacts: (i) an interaction with the RNA polymerase α subunit C‐terminal domain (αCTD) that facilitates initial binding of RNA polymerase to promoter DNA; and (ii) an interaction with the RNA polymerase α subunit N‐terminal domain that facilitates subsequent promoter opening. We have used random mutagenesis and alanine scanning to identify determinants within αCTD for transcription activation at a Class II CRP‐dependent promoter. Our results indicate that Class II CRP‐dependent transcription requires the side chains of residues 265, 271, 285–288 and 317. Residues 285–288 and 317 comprise a discrete 20×10 Å surface on αCTD, and substitutions within this determinant reduce or eliminate cooperative interactions between α subunits and CRP, but do not affect DNA binding by α subunits. We propose that, in the ternary complex of RNA polymerase, CRP and a Class II CRP‐dependent promoter, this determinant in αCTD interacts directly with CRP, and is distinct from and on the opposite face to the proposed determinant for αCTD–CRP interaction in Class I CRP‐dependent transcription.
Transcription initiation requires coordinated protein–protein and protein–DNA interactions, frequently involving one or more transcription factors in addition to the subunits of RNA polymerase. Escherichia coli holo RNA polymerase (RNAP) is a complex of four different subunits, with subunit composition α2ββ′σ, and is capable of interacting with at least three different promoter elements and with a wide range of protein transcription factors. Transcription factors are known to interact with each of the four RNAP subunits (reviewed in Rhodius and Busby, 1998), but the most studied target is the C‐terminal domain of the α subunit (αCTD). Each RNAP α subunit consists of two independent domains joined by a flexible linker. The C‐terminal domain (residues 235–329) functions both as a sequence‐specific DNA‐binding protein (recognizing ‘UP elements’ within some promoters; Ross et al., 1993) and as the target for a number of transcription activator proteins (reviewed by Ebright and Busby, 1995). The surface of αCTD that is involved in sequence‐specific interaction with DNA has been identified and characterized (Jeon et al., 1995; Gaal et al., 1996; Murakami et al., 1996). This work is concerned with the identification of residues within αCTD that are required for transcription activation by the cyclic AMP receptor protein (CRP; also known as the catabolite activator protein, CAP), a well‐characterized dimeric transcription factor that regulates the expression of >100 genes in response to changes in the intracellular concentration of its allosteric effector, cAMP (reviewed by Kolb et al., 1993).
Recent studies have suggested that any specific interaction between RNAP and a DNA‐binding protein may be sufficient to recruit RNAP to a promoter and increase the rate of transcription initiation (Dove et al., 1997). Activators that function by recruiting RNAP may therefore require nothing more of their ‘targets’ than proximity to the activator‐binding site and a particular array of surface‐exposed side chains, and it is possible that an activating region of such a transcription factor might contact different targets at promoters with different architectures. Simple CRP‐dependent promoters can be grouped into two classes depending on the location of the 22 bp DNA site for CRP (Ebright, 1993; Busby and Ebright, 1997). At Class I CRP‐dependent promoters, a single CRP dimer binds upstream of RNAP, at sites centred close to positions −61, −71, −82 or −92. At Class II CRP‐dependent promoters, a single CRP dimer binds close to position −41, overlapping the −35 element, and αCTD binds to the DNA upstream of the CRP dimer (Attey et al., 1994; Belyaeva et al., 1996; Murakami et al., 1997). At both Class I and Class II CRP‐dependent promoters, CRP activates transcription by making direct contacts with RNA polymerase. At Class I CRP‐dependent promoters, CRP makes a single interaction between a surface‐exposed loop (residues 156–164; activating region 1, AR1) in the downstream subunit of the CRP dimer and αCTD (Zhou et al., 1993; reviewed by Ebright, 1993). At Class II CRP‐dependent promoters, CRP makes two contacts with the RNAP α subunit: AR1 of the upstream subunit of the CRP dimer contacts αCTD (Zhou et al., 1994b), and a surface consisting of residues 19, 21, 96 and 101 (activating region 2, AR2) on the downstream subunit of the CRP dimer contacts the α subunit N‐terminal domain (Niu et al., 1996; reviewed by Busby and Ebright, 1997). At both classes of promoters, the CRP–αCTD interaction can be disrupted by substitutions within AR1 of CRP (Bell et al., 1990; Zhou et al., 1994a). However, the quantitative effects of individual alanine substitutions within this loop are different at Class I and Class II CRP‐dependent promoters, suggesting that the ‘targets’ in αCTD contacted by AR1 at the two classes of promoter may differ (Zhou et al., 1994a).
Previous studies have identified single amino acid substitutions within αCTD that interfere with activation by CRP at the Class I lac P1 promoter. These substitutions fall within the DNA‐binding surface, i.e. residues 265–269 and 296–299 (Zou et al., 1992; Tang et al., 1994; Gaal et al., 1996; Murakami et al., 1996), and also within an adjacent but distinct set of residues, i.e. residues 258–261 (Tang et al., 1994). Niu et al. (1996) identified mutant α subunits that were specifically defective in Class II CRP‐dependent transcription (i.e. defective in Class II CRP‐dependent transcription but not in Class I CRP‐dependent transcription or CRP‐independent transcription). Their screen defined a determinant in the α subunit N‐terminal domain, which was shown to interact with AR2 of CRP, but the determinant which interacts with AR1 was not defined.
In this work, we have used both random mutagenesis and systematic alanine scanning to identify the residues within αCTD that are required for activation by CRP at a Class II CRP‐dependent promoter. Since at these promoters CRP makes multiple contacts with RNAP, the importance of the AR1–αCTD interaction varies from one promoter to another (Rhodius et al., 1997). To ensure that substitutions which disrupted this interaction would have a measurable effect, we used the semi‐synthetic CC(−41.5) promoter that contains a consensus CRP‐binding site centred at position −41.5 upstream of the melR core promoter elements. Wild‐type CRP increases expression from this promoter ∼40‐fold, and >90% of this activation is lost if either AR1 or AR2 of CRP is disrupted (Bell et al., 1990; Niu et al., 1996; Rhodius et al., 1997). Our results indicate that transcription activation at Class II promoters requires residues within a new determinant that is distinct from the determinants known to be required for transcription activation at Class I CRP‐dependent promoters.
Random mutagenesis screen for substitutions that disrupt activation by CRP at CC(−41.5)
To identify substitutions within αCTD that interfere with transcription activation at a Class II CRP‐dependent promoter, we performed random mutagenesis of codons 231–329 of the rpoA gene encoding the α subunit of RNAP and screened for mutations that decreased the level of expression from a CC(−41.5)::lac operon fusion in vivo. The region of rpoA that encodes residues 231–329 of the α subunit was amplified by error‐prone PCR, and the mutagenized fragments were cloned into an rpoA expression plasmid (pLAW2). The resulting library of mutagenized plasmids was transformed into M182 Δlac Δcrp cells that had been transformed with plasmid pDW300 (encoding CRP) and a plasmid carrying a CC(−41.5)::lac operon fusion [pRW50/CC(−41.5)]. The transformed cells therefore contained wild‐type α subunits expressed from the chromosomal rpoA gene and mutant α subunits expressed from the multicopy plasmid‐borne rpoA gene. The transformants were plated onto lactose‐MacConkey indicator agar: colonies containing the wild‐type pLAW2 plasmid exhibited a Lac+ phenotype (red colonies). We reasoned that colonies containing pLAW2 derivatives that expressed α subunits that interfered with expression from the CC(−41.5) promoter would exhibit a Lac− phenotype (white or light red colonies; cf. Zou et al., 1992; Tang et al., 1994; Gaal et al., 1996; Niu et al., 1996).
We performed four independent mutagenesis reactions, screened ∼6000 colonies and isolated 26 independent mutants. The sequence of the mutagenized region of each candidate plasmid was determined, and the amino acid substitutions that interfered with expression from the CC(−41.5) promoter were inferred: 20 of the mutants contained single amino acid substitutions (Table I; Figure 1A). Six of the 14 different residues at which substitutions were isolated form a discrete surface‐exposed region that has not been implicated previously in DNA binding or transcription activation by CRP (residues 285, 286, 287, 288, 289 and 315). Five of the residues at which substitutions occur are important components of the DNA‐binding surface of αCTD (residues 265, 268, 296, 298 and 299; Gaal et al., 1996; Murakami et al., 1996), and two further residues (294 and 302) are adjacent to the DNA‐binding surface. The final substitution, EG261, occurs at a residue that previously has been shown to be important for transcription activation by CRP at the lac P1 promoter (Tang et al., 1994). Interestingly, in a similar screen using a CC(−41.5)::lacZ fusion, Niu and Ebright identified four substitutions in αCTD that gave weak phenotypic changes on Lac‐tetrazolium indicator plates: RC317 (part of the surface defined by residues 285–289 and 315; Figure 1A), DA258, DG259 and DN259 (part of the surface defined by residue 261; Figure 1A) (W.Niu and R.H.Ebright, unpublished data).
To quantify the effects of the substitutions listed in Table I, the pLAW2 derivatives were transformed into a reporter strain carrying a single copy chromosomal CC(−41.5)::lacZ fusion, and β‐galactosidase activity was measured. Substitutions at positions 285, 286, 287, 288 and 315 had the greatest effect, decreasing expression to between 53 and 71% of the wild‐type level. Some substitutions in the DNA‐binding surface (at residues 265, 296 and 299) also decreased expression (Table I). Curiously, a number of the substitutions that gave rise to a Lac− phenotype on indicator plates did not decrease the level of β‐galactosidase expression measured in this direct assay. The plate phenotypes may have been misleading, or the effect of these substitutions may be greater in the growth conditions prevailing during growth on plates than in the fast growing, well‐aerated cultures used for the assay.
Alanine scan to identify residues within αCTD that are important for activation by CRP at CC(−41.5)
A complication of the random mutagenesis procedure is that the substitutions obtained may confer their phenotype by introducing a clash, rather than by removing an essential side chain. Furthermore, the randomly generated library of mutants screened was not saturated, i.e. we did not carry out the screen in a manner that would have resulted in our obtaining every potential substitution that would cause a reduction in CRP‐dependent activation at CC(−41.5). To determine the importance of each side chain within αCTD for Class II CRP‐dependent transcription, we therefore performed alanine scanning (Cunningham and Wells, 1989). Lysogens carrying a chromosomal CC(−41.5)::lacZ fusion were transformed with a set of plasmids encoding the RNAP α subunit in which residues 255–329 were each changed individually to alanine, and the level of lacZ expression was determined in vivo from β‐galactosidase assays.
Alanine substitution of residues 257, 258, 265, 271, 285–288, 317 and 318 resulted in defects in Class II CRP‐dependent transcription, with substitution of residues 285–288 and 317 causing the largest defects (Figure 2). Consistent with the location of the randomly generated substitutions isolated in the earlier screen, these results suggest that there may be three discrete determinants for Class II CRP‐dependent transcription on the surface of αCTD: residues 285–289, 315, 317 and 318 form a contiguous patch on the surface of αCTD, residues 265, 268, 294, 296, 298, 299 and 302 correspond closely to the DNA‐binding surface (Gaal et al., 1996; Murakami et al., 1996), and residues 257–259, 261 and 271 form another contiguous patch on the opposite face of the domain to residues 285–289, 315, 317 and 318 (Table I; Figures 1 and 2).
In vitro transcription at a Class II CRP‐dependent promoter
The reporter strains used in the experiments described above contained wild‐type α subunits that may have diminished the observed effects of the mutant α subunits. Because of this, and the other complications inherent in interpreting results obtained from a multicomponent system in vivo, we complemented our in vivo analyses with in vitro transcription experiments, using purified RNAP reconstituted with α subunits carrying selected alanine substitutions. His‐tagged α subunits carrying alanine substitutions at positions 258, 261, 265, 271, 285, 286, 287, 288 or 317 were overexpressed, purified and reconstituted into RNAP with wild‐type β, β′ and σ70 subunits [although substitution of residue 261 had no effect in the alanine scan performed at CC(−41.5) in vivo, αEA261 was included because of the reported importance of residue 261 at the Class I CRP‐dependent lac P1 promoter; Tang et al. (1994)]. The activity of each RNAP preparation was normalized to transcription from a template carrying the constitutive lacUV5 promoter. The ability of each RNAP to transcribe from a template carrying the CC(−41.5) promoter, in the absence or presence of wild‐type CRP, was then measured.
None of the RNAPs produced a detectable level of transcript from the CC(−41.5) promoter in the absence of CRP. Transcription by each RNAP was stimulated by the presence of wild‐type CRP, but the level of CRP‐dependent transcription differed greatly between the mutant RNAPs (Figures 3 and 4). The greatest defects in CRP‐dependent transcription were found with RNAPs carrying αTA285 or αVA287 subunits, which produced ∼20% of the amount of CRP‐dependent transcript produced by wild‐type RNAP. Alanine substitutions at positions 265, 271, 286, 288 and 317, but not 258 or 261, also resulted in defects in Class II CRP‐dependent transcription in this assay. The in vitro transcription experiments demonstrate that alanine substitution of residues 265, 271, 285–288 and 317 of the α subunit directly affects the level of transcription from the CC(−41.5) promoter, and exclude the possibility that the phenotypes observed in vivo could have resulted from indirect effects, such as altered levels of CRP expression. We conclude that, for residues 265, 271, 285–288 and 317, side chain atoms beyond Cβ make favourable interactions in Class II CRP‐dependent transcription.
Transcription from CC(−41.5) was also measured in the presence of CRP carrying a Leu for His substitution at position 159, which completely inactivates AR1 (Rhodius et al., 1997). Each RNAP produced a low, but detectable, level of transcript in the presence of this mutant activator protein (∼10% of the level produced by wild‐type RNAP in the presence of wild‐type CRP; Figure 3; data not shown). As the HL159 substitution in CRP causes a greater defect in transcription activation than any of the single alanine substitutions in αCTD, it appears that more than one side chain must be removed from αCTD in order to destroy the AR1–αCTD interaction completely.
Interactions between the α subunit of RNAP and DNA
To determine if any of the alanine substitutions within αCTD that interfered with transcription activation by CRP also caused defects in sequence‐specific interactions between the α subunit of RNAP and DNA, we measured transcription at the UP element‐dependent rrnB P1 promoter in vitro. Templates carrying rrnB P1 promoter derivatives with or without a UP element were transcribed by RNAPs reconstituted with α subunits carrying single alanine substitutions at positions 261, 265, 285–288, 317 (Figure 5) or 271 (data not shown). Consistent with previous reports (Gaal et al., 1996; Murakami et al., 1996), of the substitutions tested here, only RA265 severely reduced transcription from the rrnB P1 promoter containing a UP element. We conclude that the defects in CRP‐dependent transcription caused by the other substitutions cannot be ascribed to altered interactions with DNA.
Interactions between the α subunit of RNAP and CRP
We previously described a series of semi‐synthetic promoters that carry the rrnB P1 UP element at different locations adjacent to a consensus CRP site, and showed that, at some of these promoters, purified α subunits and CRP bind cooperatively. The observed cooperativity is dependent on AR1 of CRP, and is destroyed if CRP carries the HL159 substitution, suggesting that the cooperativity directly reflects α subunit–CRP interactions (Savery et al., 1995; Lloyd et al., 1998). Figure 6 shows the results of electrophoretic mobility shift experiments with the CC(−41.5) α(−63) promoter, a derivative of CC(−41.5) in which the rrnB P1 UP element is located immediately upstream of the CRP‐binding site (Lloyd et al., 1998). Purified His‐tagged wild‐type α subunits and αEA261 subunits exhibit ∼4‐fold cooperativity with wild‐type CRP. In contrast, purified His‐tagged αVA287 and αGD315 subunits exhibited no cooperativity with CRP in interacting with CC(−41.5) α(−63). Thus, the VA287 and GD315 substitutions reduce α subunit–CRP interactions.
We have identified seven residues within αCTD whose side chains are important for transcription activation by CRP at CC(−41.5) both in vivo and in vitro (R265, K271, T285, E286, V287, E288, R317). The location of these residues, together with the location of non‐alanine substitutions that affect expression from CC(−41.5) in vivo, suggests that αCTD contains at least two, and possibly three, separate determinants for transcription activation at Class II CRP‐dependent promoters (Figure 1).
The determinant that has the greatest effect on transcription activation by CRP at CC(−41.5), both in vivo and in vitro, includes residues 285–289, 315, 317 and 318. This determinant is not required for α subunit–DNA interaction but is required for cooperative α subunit–CRP interactions (Figures 5 and 6). We propose that this surface of αCTD makes direct protein–protein interactions with AR1 of CRP in the ternary complex of RNAP, CRP and a Class II CRP‐dependent promoter. Consistent with this proposal, the dimensions of this determinant (∼20×10 Å; Figure 1; Jeon et al., 1995) are comparable with the dimensions of AR1 (∼14×11 Å; Niu et al., 1994), and the side chain identities of the two most critical residues within the determinant, T285 and V287, suggest possible hydrogen‐bonded and hydrophobic interactions with the side chains of the two most critical residues of AR1, T158 and P160 (Zhou et al., 1994a). Interestingly, there are precedents for the involvement of this region of αCTD in transcription activation: substitutions at residues 289 and 317 decreased transcription activation by the CRP homologue FNR in vivo (Lombardo et al., 1991), and substitutions at 286, 287, 289, 290 and 300 decreased transcription activation by bacteriophage P2 Ogr in vivo (Wood et al., 1997).
The second determinant includes residues 265, 268, 294, 296, 298, 299 and 302, which correspond closely to the DNA‐binding surface of αCTD (Jeon et al., 1995; Gaal et al., 1996; Murakami et al., 1996). We propose that this determinant makes non‐specific protein–DNA interactions with the DNA segment adjacent to the CRP site in the ternary complex of RNAP, CRP and a Class II CRP‐dependent promoter. This is analogous to the proposed role of this determinant in Class I CRP‐dependent transcription (Blatter et al., 1994; Busby and Ebright, 1994; Gaal et al., 1996), Mor‐dependent transcription (Artsimovitch et al., 1996) and Ogr‐dependent transcription (Wood et al., 1997). Consistent with this proposal, results from DNase I footprinting (Attey et al., 1994), hydroxyl radical footprinting (Belyaeva et al., 1996) and experiments with a chemical DNA‐cleaving reagent (Murakami et al., 1997) indicate that αCTD is close to, or in contact with, the DNA segment immediately upstream of the CRP site in the ternary complex of RNAP, CRP and a Class II CRP‐dependent promoter. In addition, substitution of this region of the CC(−41.5) promoter with the UP element from rrnB P1 increases promoter strength (Savery et al., 1995; Lloyd et al., 1998).
Random mutagenesis and alanine scanning implicate residues 257, 258, 259, 261 and 271 as a potential third determinant for Class II CRP‐dependent transcription (Table I; Figure 2; W.Niu and R.H.Ebright, unpublished data). These residues form a discrete surface‐exposed patch that is on the opposite face of αCTD to the 285–289, 315, 317, 318 determinant (Figure 1). Residues within this determinant have been proposed to make direct interactions with AR1 of CRP at the Class I CRP‐dependent lac promoter (Tang et al., 1994). However, the role and relative importance of these residues at the Class II CRP‐dependent promoter studied here is unclear; the effects of substitutions in this determinant are small in vivo and, except for substitution of residue 271, not detectable in vitro (Table I; Figures 2). It is possible that these substitutions affect interactions other than the αCTD–CRP and αCTD–DNA contacts.
One interpretation of our results is that the face of αCTD that interacts with AR1 of CRP at Class II CRP‐dependent promoters differs from the face that interacts at Class I CRP‐dependent promoters. An alternative interesting possibility is that, at any CRP‐dependent promoter, αCTD is able to interact with DNA in either of two orientations; one that presents the 285–289, 315, 317, 318 determinant to CRP, and one that presents the 257–259, 261, 271 determinant to CRP, with the preferred orientation depending on the architecture of the promoter and the sequence of the DNA segment contacted by αCTD. Consistent with this, we have observed that substitutions in the 285–289, 315, 317, 318 determinant, in addition to substitutions in the 257–259, 261, 271 determinant, reduce transcription at Class I CRP‐dependent promoters (N.J. Savery, S.J.W.Busby and R.L.Gourse, unpublished data; W.Niu and R.H.Ebright, unpublished data). Further work is now essential to elucidate the variables affecting the relative utilization of the two patches at Class I and Class II CRP‐dependent promoters.
Materials and methods
Strains and plasmids
Table II lists the strains and plasmids used in this work. Standard methods for isolation and manipulation of DNA fragments were used throughout. Strain RLG4649 [a λ lysogen encoding a chromosomal CC(−41.5)::lacZ fusion] was constructed from strain M182 by the method of Simons et al. (1987), as described in Rao et al. (1994). Plasmid pSR/CC(−41.5) was constructed by cloning an EcoRI–HindIII fragment carrying the CC(−41.5) promoter (Gaston et al., 1990) into plasmid pSR (Kolb et al., 1995). The promoter fragment used for the construction of both strain RLG4649 and plasmid pSR/CC(−41.5) carried the CC(−41.5) DNA sequence from position −80 upstream of the transcription start point to position +35 downstream. Plasmid pSR/lacUV5 was constructed by cloning an EcoRI–HindIII fragment, carrying lacUV5 promoter sequence from −60 to +40, from RLG593 (Ross et al., 1990) into plasmid pSR.
Random mutagenesis of the αCTD coding region
Error‐prone PCR (Zhou et al., 1991) was used to prepare a library of random rpoA mutations. The rpoA αCTD coding region from plasmid pLAW2 was amplified by four independent PCRs using Taq DNA polymerase and oligonucleotide primers that flanked the HindIII site adjacent to codon 231 and the BamHI site downstream of the translation stop codon. After restriction with HindIII and BamHI, the products were cloned into pLAW2 to generate a library of pLAW2 derivatives carrying random mutations in the segment encoding αCTD. M182 Δcrp cells carrying pDW300, encoding CRP, and pRW50/CC(−41.5), encoding a CC(−41.5)::lac fusion, were transformed by electroporation with samples from the pLAW2 library, and the transformed cells were plated onto MacConkey indicator agar containing 10 g/l lactose, 100 μg/ml ampicillin, 35 μg/ml tetracycline and 25 μg/ml kanamycin (this reporter strain was used because it gives a stronger Lac+ phenotype on the indicator plates than a reporter strain containing a chromosomal crp gene).
Measurement of β‐galactosidase activity in vivo
Cultures were inoculated to an A600 of ∼0.007 and grown to mid‐log phase (A600 ∼0.35–0.40) in Lennox Broth containing 100 μg/ml ampicillin at 37°C with vigorous aeration. Cultures of strains transformed with derivatives of plasmid pLAW2 also contained 1 mM isopropyl‐β‐d‐galactopyranoside (IPTG). Specific β‐galactosidase activity was determined by the method of Miller (1972).
Protein purification and reconstitution of RNA polymerase
RNAP α subunits carrying a hexa‐His tag between the first and second codons were prepared using plasmid pHTT7f1NHα. Derivatives of pHTT7f1NHα carrying mutant rpoA alleles were constructed by replacing the HindIII–BamHI fragment, which encodes αCTD, with fragments from plasmids encoding the different amino acid substitutions (Tang et al., 1994; Gaal et al., 1996; Wood et al., 1997; M.Kainz and R.L.Gourse, in preparation). Overexpression of the α subunits in strain BL21 DE3 and purification of α by Ni2+‐affinity chromatography were performed essentially as described in Tang et al. (1995) and Gaal et al. (1996). Preparation of inclusion bodies containing β, β′ or σ70 subunits from strains XL1‐Blue [pMKSe2], BL21 DE3 [pT7β′] and BL21 DE3 [pLHN12σ], respectively, and reconstitution of RNA polymerase were also performed as described in Tang et al. (1995). Wild‐type CRP and CRP HL159 were purified by the method of Ghosaini et al. (1988) from M182 Δcrp cells transformed with plasmids pDCRP or pDCRP HL159.
In vitro transcription assays
In vitro transcription reactions (25 μl) were performed in a buffer containing 100 mM KCl, 40 mM Tris–acetate pH 7.9, 10 mM MgCl2, 1 mM dithiothreitol (DTT), 100 μg/ml bovine serum albumin (BSA), 200 μM ATP, 200 μM CTP, 200 μM GTP, 10 μM UTP and 5 μCi of [α‐32P]UTP (DuPont). CRP, if added, was present at 20 nM. Reactions that contained CRP also contained 0.2 mM cAMP. Supercoiled template DNA was prepared using a Qiagen midiprep kit, and was added at a final concentration of 0.2 nM. Reactions were started by the addition of RNAP and incubated at 22°C for 15 min. Samples were analysed by denaturing gel electrophoresis and quantified using a PhosphorImager (Molecular Dynamics) and ImageQuant software. Concentrations of RNAP, chosen to give the same amount of transcription from the constitutive lacUV5 promoter in the absence of CRP, were 2.4 nM wild‐type RNAP, 1.8 nM RNAP αDA258, 2.2 nM RNAP αEA261, 12.5 nM RNAP αRA265, 2.3 nM RNAP αKA271, 6.7 nM RNAP αTA285, 4.7 nM RNAP αEA286, 2.6 nM RNAP αVA287, 5.9 nM RNAP αEA288 and 5.9 nM RNAP αRA317.
Electrophoretic mobility shift assays
Electrophoretic mobility shift assays were performed with EcoRI–HindIII fragments prepared from plasmid pAA121/CC(−41.5) α(−63) and end‐labelled with [γ‐32P]ATP and polynucleotide kinase. Reaction mixtures contained ∼0.4 nM labelled DNA, 20 mM HEPES pH 8.0, 5 mM MgCl2, 50 mM potassium glutamate, 1 mM DTT, 2 mM spermidine, 20 μg/ml sonicated herring sperm DNA, 10% glycerol and proteins as indicated in the figure legends. Reaction mixtures that contained CRP also contained 0.2 mM cAMP. Reactions were incubated for 20 min at 37°C and were then run on 6% acrylamide gels containing 7.5% glycerol and 0.5× TBE buffer. Gel running buffer was 0.5× TBE containing 2% glycerol.
We thank Gail Christie for providing us with plasmids encoding α subunits carrying alanine substitutions at positions 288–290. This work was supported by project grant G07974 from the UK BBSRC to S.J.W.B., USPHS grant GM37048 from the National Institutes of Health to R.L.G., grant GM41376 from the National Institutes of Health and an HHMI Investigatorship to R.H.E., and by a Human Frontier Science Program fellowship to N.J.S.
- Copyright © 1998 European Molecular Biology Organization