INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Regulation of the Escherichia coli rhaBAD operon responds to both availability of L-rhamnose and catabolite repression. In the presence of L-rhamnose, the AraC family activator RhaS (reviewed in reference 12) binds to a site that spans from position 32 to position 81 relative to the rhaBAD transcription start site (9, 10). This RhaS-binding site consists of two 17-bp inverted repeat half-sites that are separated by 16 bp of DNA not contacted by RhaS (9). RhaS alone can activate rhaBAD expression approximately 1,000-fold above the extremely low basal level (10). The cyclic AMP receptor protein (CRP) mediates catabolite repression at rhaBAD by binding to a site immediately upstream of RhaS that is centered at position 92.5 relative to the rhaBAD transcription start site (10). CRP alone does not activate rhaBAD expression, but in the presence of RhaS CRP can contribute 30- to 50-fold additional activation (10).

CRP is a global regulator of catabolite repression in E. coli (reviewed in reference 6). Interactions between CRP and RNA polymerase (RNAP) that are required for transcription activation have been well defined for promoters where CRP is the only activator. These simple CRP-dependent promoters are categorized according to the location of the CRP-binding site. At class I CRP-dependent promoters CRP binds upstream but not adjacent to RNAP, with sites for CRP usually centered at positions 62.5, 72.5, or 92.5 relative to the transcription start site. CRP activation at class I promoters involves protein-protein contacts between a surface-exposed loop on CRP activating region 1 (AR1), and the carboxyl-terminal domain of the  subunit (-CTD) of RNAP (31, 35, 36; reviewed in reference 6). At class II CRP-dependent promoters CRP binds to a site that is centered at position 42.5 and overlaps the 35 region. In this situation, contacts are made between a second activating region on CRP, AR2, and the N-terminal domain of  (-NTD) (21, 27, 29; reviewed in references 5 and 6), as well as between CRP AR1 and -CTD (32, 36).

Activation by CRP at promoters where CRP acts in conjunction with a regulon-specific activator, called class III promoters, has been less thoroughly studied. In contrast to class I and class II promoters, a pattern or patterns for the role of CRP at class III promoters has not yet emerged. For example, at the uhpT promoter, CRP binds at position 103.5 and acts in conjunction with the uhp-specific activator, UhpA, bound at position 64. Merkel et al. (17) found that CRP AR1 substitutions were not defective for uhpT activation, suggesting that CRP activation of uhpT does not depend on the previously defined -CTD-AR1 interactions. More recent work has shown that -CTD is required for uhpT activation (23). CRP is also involved in activation with regulon-specific proteins at several pairs of divergent promoters. At some divergent promoters, such as mal, the only role of CRP appears to be to reposition other activators (28), while at others there is evidence that CRP plays a role in both DNA structure and in protein-protein interactions (4, 25, 26, 34).

-CTD is dispensable at many activator-independent promoters; however, it is required for interaction with DNA (especially UP elements) or activator proteins and DNA at a large number of other promoters. Three determinants on the surface of -CTD have been identified based on their functions at class I CRP-dependent promoters and UP elements (reviewed in reference 6). The 265 determinant is important for both UP element activation and CRP-dependent activation and is proposed to identify residues involved in DNA binding. The 261 determinant of -CTD is also important for both UP element and CRP-dependent activation; however, these residues are not required for -CTD interactions with DNA, and their function is not clear. The 287 determinant is required for activation by CRP but is not required for DNA binding or UP element-dependent activation. Mutations in the 287 determinant disrupt cooperativity between CRP and ; hence, it is proposed that these residues are required for interactions between CRP and -CTD.

The goal of this work was to characterize rhaBAD transcription activation by CRP and -CTD. We found that alanine substitution of some residues within both AR1 and AR2 of CRP resulted in small defects in rhaBAD activation. To determine whether -CTD was required for rhaBAD activation, we expressed a derivative of  deleted for the entire C-terminal domain, -235. Expression of -235 resulted in a 54-fold defect at a promoter with only a RhaS-binding site and a 13-fold defect at a promoter with binding sites for both CRP and RhaS. Deletion of rhaS from the cell eliminated the -CTD deletion defects at all promoters. Using a library of alanine substitutions in -CTD, we found strong evidence for an -CTD interaction with DNA, suggestive evidence for a possible interaction between -CTD and RhaS, and no evidence for an -CTD-CRP interaction. Overall, our results are most consistent with a model for rhaBAD activation in which CRP activates by a mechanism other than interaction with -CTD and in which -CTD activates by interacting with DNA and possibly RhaS.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

General methods. Standard methods were used for restriction endonuclease digestion, ligation, and transformation of DNA. Most DNA sequences were verified using automated dideoxy sequencing on a LI-COR 4000L sequencer. Primers for the LI-COR 4000L were labeled with IRD-41 and were custom made by LI-COR, Inc. (Lincoln, Nebr.). The Thermo Sequenase fluorescent-labeled primer cycle sequencing kit from Amersham Life Science (Piscataway, N.J.) was used for sequencing reactions. Additional DNA sequence verification was performed on an ABI Prism 310. Primers for ABI Prism sequencing were synthesized by Oligos, Etc. (Wilsonville, Oreg.), and sequencing reactions were carried out using a Thermo Sequenase dye terminator sequencing kit from Amersham Life Science.

Culture media. Morpholinepropanesulfonic acid (MOPS; 1×) buffered medium (20) was used to grow cultures for -galactosidase assay and consisted of 40 mM MOPS, 4 mM Tricine, 0.01 mM FeSO₄, 9.5 mM NH₄Cl, 0.276 mM K₂SO₄, 0.5 µM CaCl₂, 0.528 mM MgCl₂, 50 mM NaCl, 3 × 10⁹ M Na₂MoO₄, 4 × 10⁷ M H₃BO₃, 3 × 10⁸ M CoCl₂, 10⁸ M CuSO₄, 8 × 10⁸ M MnCl₂, 10⁸ M ZnSO₄, 1.32 mM K₂HPO₄, 10 mM NaHCO₃, 0.2% Casamino Acids, and 0.002% thiamine. Overnight medium consisted of 1× MOPS medium containing 0.04% glycerol. Growth medium consisted of 1× MOPS medium containing 0.4% glycerol, 125 mg of ampicillin per ml, and 0.2% L-rhamnose. For other experiments (cloning, strain construction, etc.) cells were grown in tryptone-yeast (TY) medium (16) with or without antibiotic.

Plasmids, phages, and strains. The E. coli strains,  phages, and plasmids used in this study are listed in Table 2. Wild-type crp was cloned by PCR amplification using primers 2003 and 2004 (Table 1) with whole cells of strain ECL116 (Table 2) serving as template. The PCR product was digested at the EcoRI site in 2003 and the BamHI site in 2004 and cloned between the EcoRI and BamHI sites of pHG165, resulting in pSE186. The DNA sequence of the entire cloned crp gene was verified by automated sequencing on both strands.

-Galactosidase assay. Strains for -galactosidase assay were grown as described by Bhende and Egan (2). Briefly, starter cultures were grown in tryptone-yeast extract broth (with 125 µg of ampicillin per ml added for strains containing plasmid) for approximately 7 h at 37°C. Then, 40 µl of starter culture was used to inoculate 2.5 ml of 1× MOPS overnight medium (see recipe above), and this was grown for approximately 17 h. An approximately 200-µl volume of the overnight culture, or the appropriate volume to reach a starting optical density at 600 nm of 0.01 to 0.02 in the growth flask, was then used to inoculate 10 ml of 1× MOPS growth medium (see recipe above) in 125-ml baffled flasks. Cultures were grown at 37°C with vigorous shaking (~280 rpm) to an A₆₀₀ of approximately 0.4. After resuspension of the cell pellets in Z buffer (18), -galactosidase activity was determined as described by Miller (18) except that incubation with substrate (o-nitrophenyl--D-thiogalactopyranoside) was done at room temperature. Specific activities were averaged from at least three independent assays, with two replicates in each assay.

Mutagenesis of crp. Oligonucleotide primers for site-directed mutagenesis (Table 1) were synthesized by NBI (Beverly, Mass.) and Oligos, Etc. Alanine substitutions were introduced using the Gene Editor In Vitro Mutagenesis System (Promega Corp., Madison, Wis.) with plasmid pSE186 as template. In all cases, the entire crp gene was sequenced on both strands to confirm the mutation and to confirm that there were no additional mutations.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

CRP AR1 and AR2 substitution derivatives at rhaBAD. To begin to understand the mechanism of CRP activation of rhaBAD expression, we wished to determine whether AR1 and/or AR2 of CRP are necessary for activation at rhaBAD. We tested two alanine substitutions in AR1 (T158A and G162A) and three in AR2 (H19A, H21A, and K101A) which cause approximately 5- to 40-fold activation defects at standard class I and II promoters (21, 22, 35). Each of the CRP substitution derivatives was tested at two different fusions of the rhaBAD promoter region with lacZ. In the first fusion [(rhaB-lacZ)226; Fig. 1], rhaBAD promoter DNA extended upstream to position 226 and included the RhaS and CRP-binding sites at rhaBAD as well as the RhaR and CRP-binding sites at rhaSR. At (rhaB-lacZ)226, G162A and H21A both resulted in approximately twofold defects, and K101A was also slightly defective (Table 3). In the second fusion [(rhaB-lacZ)110], rhaBAD promoter DNA extended upstream to position 110 and included only the RhaS and CRP-binding sites at rhaBAD. At (rhaB-lacZ)110, G162A and H21A again resulted in approximately twofold defects, K101A was again slightly defective, and H19A also resulted in nearly a twofold defect.

View larger version (25K): [in this window] [in a new window]   FIG. 1.   rhaSR-rhaBAD intergenic region. The top line shows a schematic representation of the regulatory region between rhaBAD and rhaSR. The relative positions of the two RNA polymerases and the activator proteins RhaS, CRP, and RhaR are shown. The bottom lines show the DNA sequence upstream of rhaBAD extending back to position 239 (the rhaSR transcription start site). The positions of the RhaS and RhaR-binding sites are shown by everted arrows, and the position of the CRP-binding site is shown by inverted arrows. The 10 and 35 regions of the two promoters are marked. The upstream endpoints of rhaBAD promoter fusions are identified.

View this table: [in this window] [in a new window]   TABLE 3.   Effects of alanine substitutions in CRP AR1 and AR2 at (rhaB-lacZ)

View this table: [in this window] [in a new window]   TABLE 4.   Effects of alanine substitutions in CRP AR1 and AR2 at (rhaB-lacZ)84

Effect of deleting -CTD at rhaBAD. To determine whether -CTD is important for activation of rhaBAD, we constructed a plasmid that expresses a derivative of  deleted for the entire -CTD, -235. We assayed (rhaB-lacZ) expression in a strain expressing wild-type  or -235 from a plasmid as well as wild-type  from the chromosome (Table 5). At the shortest promoter, (rhaB-lacZ)70, there was a sixfold defect upon expression of -235. There was no difference between expression of wild-type  and -235 at the same promoter in a rhaS strain background, indicating that activation by -CTD requires RhaS. It is likely that the higher level of expression (0.056U) from (rhaB-lacZ)70 in the rhaS⁺ strain expressing wild-type  reflects some ability of RhaS to bind to the partial RhaS-binding site that remains in this fusion, at least in the presence of -CTD.

View this table: [in this window] [in a new window]   TABLE 5.   Effect of expressing an -CTD deletion derivative at rhaBAD

RNAP -CTD alanine scan. Based upon the results of our analysis with -235, it appears that -CTD is required for full rhaBAD activation. We wished to identify residues in -CTD involved in this activation and to investigate whether -CTD activation depends on interactions with DNA and/or CRP and/or RhaS. To do this, we tested a plasmid-borne library of -CTD alanine substitution derivatives for activation at (rhaB-lacZ)110 and (rhaB-lacZ)84 in strain backgrounds that expressed wild-type  from the chromosome. The results of our analysis are shown in Fig. 2 and are summarized in Table 6. We have divided the important residues in -CTD identified at rhaBAD into "DNA-binding" and "Other" categories. Assignment of residues into the DNA-binding category was based on the residues identified as part of the DNA-binding "265 determinant" based on studies with CRP and UP elements (6, 11, 19), on the fluorescence characterization of -CTD binding to a factor-independent promoter and CRP and UP element-dependent promoters (24), and on the predicted position of the residues on the structure of -CTD (accession no. 1COO.PDB).

View larger version (77K): [in this window] [in a new window]   FIG. 2.   Effect of single alanine substitutions within -CTD on activation from (rhaB-lacZ)110 (SME1035) and (rhaB-lacZ)84 (SME1036). Activities are expressed as a percentage of the average activity measured from cells transformed with plasmids encoding the wild-type  subunit. Values shown are the averages of at least three independent experiments. Each -CTD alanine substitution that significantly lowered expression compared to wild-type -CTD as determined by analysis of variance statistical analysis is indicated by an asterisk above the bar.

View this table: [in this window] [in a new window]   TABLE 6.   Summary of important residues in -CTD

View larger version (68K): [in this window] [in a new window]   FIG. 3.   Space-filling model of predicted -CTD structure. The model was based on the atomic coordinates of Jeon et al. (15). Colored residues are those identified as important at the (rhaB-lacZ)84 promoter fusion. Orange residues are those that may be involved in interaction with DNA (DNA-binding category), while violet residues are the residues that are unlikely to be involved in interaction with DNA (Other category). Residue numbers for some of the important residues are shown. The two models are related to one another by a 90° rotation around the vertical axis.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Model for activation of rhaBAD expression. Our original hypothesis for the mechanism of transcription activation at the rhaBAD promoter was that RhaS activation would not require -CTD but would occur by another mechanism such as interaction with ⁷⁰. We expected that CRP activation would involve contacts with -CTD; hence, -CTD would be required for full activation. Overall, our results indicate that -CTD is required for full activation of rhaBAD and that at least some of the -CTD activation involves interaction with DNA. We have some evidence that -CTD activation, at least in the absence of CRP, may involve contacts with RhaS. CRP activation of rhaBAD expression appears to occur by a mechanism other than interaction with -CTD.

Role of CRP AR1 and AR2 in rhaBAD expression. While substitution of some residues in CRP AR1 and AR2 resulted in small defects in rhaBAD expression, we feel that these small AR1 and AR2 defects may not indicate loss of the same interactions found at simple class I and class II CRP-dependent promoters. In fact, the defects were so small that they may simply be the result of small effects on the stability or structure of the CRP proteins. While alanine substitution of CRP T158 resulted in a 40-fold at a synthetic class I promoter (35), this substitution had no defect at rhaBAD, suggesting that AR1 may not be important at rhaBAD. The small defect upon substitution of CRP G162 leaves open the possibility that CRP activation of rhaBAD could involve AR1 contacts with -CTD; however, the evidence for such an interaction is not strong.

Is -CTD important for rhaBAD activation? The overall conclusion from our analysis of the importance of -CTD in rhaBAD activation is that some activation can occur in the absence of -CTD, but maximal activation requires the contribution of -CTD (Table 5). Two alternative mechanisms could explain this partial dependence on -CTD. First, the only contribution of -CTD to activation could be by interaction with DNA, and all of the activation signals that CRP and RhaS transmit to RNAP could be transmitted by -CTD-independent mechanisms. Alternatively, -CTD could be one of two or more sites for transmission of information from RhaS and/or CRP to RNAP.

Does activation by -CTD involve DNA contacts? Our assays identified five (positions 265, 268, 296, 298, and 299) of the seven residues within the -CTD 265 determinant which has been shown to be required for DNA binding (11, 19; reviewed in reference 6). We also found a variety of other residues that were defective at both of the tested promoters (Table 6), most of which lie very near the 265 determinant and are likely to be required for full DNA binding (Fig. 3). Hence, we conclude that -CTD contacts with DNA are important for full activation at both the full-length [(rhaB-lacZ)110] and truncated [(rhaB-lacZ)84] promoters (Table 6 and Fig. 2 and 3).

Does activation by -CTD involve contacts with RhaS? We were surprised to find that activation at a promoter including only the RhaS-binding site [(rhaB-lacZ)84] was greatly dependent on -CTD (54-fold; Table 5). This result indicates that -CTD makes productive interactions, at least with DNA, at the (rhaB-lacZ)84 promoter. However, since the defect upon expression of -235 required rhaS⁺, it is possible that -CTD also interacts with RhaS. This dependence of -CTD activation on RhaS could also be explained by other mechanisms. For example, it is possible that RhaS binding shifts -CTD to a stronger DNA site or that RNAP is capable of little to no rhaBAD transcription initiation in the absence of RhaS (recall the low rhaBAD expression in the absence of RhaS) (Table 6). In this second model, RhaS would overcome the rate-limiting step in initiation of rhaBAD transcription, and only then could -CTD contribute to activation. While (rhaB-lacZ)84 has an unnatural promoter, assays of this fusion are relevant because they may provide clues to the activation mechanism at the full-length promoter and/or insight into the mechanism of activation by AraC family proteins that function without the aid of a second activator such as CRP.

Does activation by -CTD involve contacts with CRP? As mentioned above, it was not possible to test the contribution of -CTD to CRP activation in the absence of RhaS since CRP does not activate under such conditions. Instead, we must draw conclusions about CRP activation by comparing results at the (rhaB-lacZ)84 and (rhaB-lacZ)110 fusions. Expression of -235 in a strain carrying a (rhaB-lacZ)110 fusion resulted in a smaller defect than that found at (rhaB-lacZ)84. This smaller defect may indicate that CRP displaces -CTD to a weaker DNA site, thereby reducing the ability of -CTD to contribute to activation. Alternatively, the smaller defect could indicate that the roles of -CTD and CRP in rhaBAD activation are partially redundant. Several of our results are consistent with this. First, the fold activation by -CTD increased in the absence of CRP [compare the defects upon -CTD deletion at (rhaB-lacZ)110, 13-fold, and (rhaB-lacZ)84, 54-fold]. Second, the importance of CRP increased in the absence of -CTD (compare the defects upon deleting the CRP-binding site with wild-type , 38-fold, and with expression of -235, 162-fold). This increased activation by CRP and -CTD in the absence of the other is the opposite of what would be expected if CRP activation required contacts with -CTD.