Binding of the C-Terminal Domain of the α Subunit of RNA Polymerase to the Phage Mu Middle Promoter

Abstract

The C-terminal domain of the α subunit (αCTD) of Escherichia coli RNA polymerase is often involved in transcriptional regulation. The αCTD typically stimulates transcription via interactions with promoter UP element DNA and transcriptional activators. DNase I footprinting and gel mobility shift assays were used to look for potential interaction of the αCTD with the phage Mu middle promoter P_m and its activator protein Mor. Binding of RNA polymerase to P_m in the presence of Mor resulted in production of a DNase I footprint downstream of Mor due to open complex formation and generation of a second footprint just upstream of the Mor binding site. Generation of the upstream footprint did not require open complex formation and also occurred in reactions in which the αCTD or His-α proteins were substituted for RNA polymerase. In gel mobility shift assays, the formation of a supershifted ternary complex demonstrated that Mor and His-α bind synergistically to P_m DNA. Gel shift assays with short DNA fragments demonstrated that only the Mor binding site and a single upstream αCTD binding site were required for ternary complex formation. These results suggest that the αCTD plays a role in P_m transcription by binding to P_m DNA just upstream from Mor and making protein-protein interactions with Mor that stabilize the binding of both proteins.

Transcription in prokaryotes is carried out by RNA polymerase holoenzyme, which is composed of a core enzyme (α₂ββ′ω) capable of nonspecific DNA binding and one of several σ subunits that confer promoter specificity (for reviews, see references 16, 25, 35, 36, and 40). The α subunit, which consists of two domains connected by a flexible linker (7, 37), plays important roles in polymerase assembly and transcriptional regulation. The α N-terminal domain (αNTD) contains determinants for dimer formation and serves to initiate core assembly by sequential binding of β and β′ to the α dimer (20, 52). The α C-terminal domain (αCTD) contains determinants for DNA binding and interaction with a number of transcriptional activators (10, 13, 15, 19). The binding of the αCTD to promoter DNA generally occurs upstream of the core promoter, but its precise location varies depending on both the promoter sequence and the binding of transcriptional regulatory proteins (6, 33, 34).

Typical factor-independent promoters contain −35 and −10 hexamers that are separated by a 16- to 18-nucleotide spacer region (17) and are contacted by the primary or “housekeeping” sigma factor σ⁷⁰ (9, 25). In addition to the core promoter sequences described above, strong promoters often contain an AT-rich UP element that is located upstream of the −35 hexamer and binds the αCTD (15, 41). In promoters that respond to activators, the −35 hexamer is usually poor, and activator binding to the promoter often facilitates or stabilizes RNA polymerase binding, a process called RNAP recruitment (9, 19). Other activators stimulate isomerization from a closed complex to an open complex which is capable of transcription initiation (18, 19, 32). In simple Class I activator-dependent promoters, the activator generally binds in the −60 region and contacts the αCTD, which is then bound between the activator and the remainder of RNAP (9, 19, 33, 40, 43). In simple Class II promoters, the activator generally binds in the −40 region and interacts with the C-terminal domain of σ (σCTD) or with both the σCTD and the αCTD, which is then bound upstream of the activator (9, 19, 26). At Class II promoters, the αNTD can also serve as a target for interaction with CAP, the catabolite activator protein (38).

During lytic development of bacteriophage Mu, a regulatory cascade of activator proteins controls synthesis of the middle and late transcripts by the Escherichia coli RNA polymerase holoenzyme containing σ70 (14, 27-31, 51). Whereas the early promoter P_e is recognized directly by the host RNAP (22), the middle promoter P_m and late promoters P_lys, P_I, P_P, and P_mom lack recognizable −35 hexamers (28, 30, 47) and require activation by the Mu Mor and C proteins, respectively (28-30).

The 129-amino-acid Mor protein forms an intertwined symmetrical dimer (23) and binds to an imperfect inverted-repeat dyad-symmetry element centered at −43.5 in P_m (2, 30). Activation of transcription from P_m by Mor requires both the σCTD and αCTD of RNA polymerase (3), making P_m a Class II activator-dependent promoter. Random mutagenesis of P_m from −70 to +10 revealed three regions important for P_m function: the −10 region, the Mor binding site from −51 to −36, and bases −31 to −29 in the spacer region, which may be involved in open complex formation (2, 4). Mutations in the region upstream of −51 had much less dramatic effects on P_m activity, but a −57 G-to-T mutation resulted in a twofold increase in activity, and a −54 A-to-T mutation reduced activity by half (1), suggesting a possible role for this region in P_m function.

Several lines of evidence suggest that the αCTD of RNAP may bind to P_m upstream of Mor and participate in transcription activation: (i) in vitro transcription experiments demonstrated that deletion of the αCTD decreased P_m activity dramatically (3); (ii) a short DNase I footprint upstream from the Mor binding site appeared only when Mor and RNA polymerase were both present (3); (iii) a yeast interaction trap assay showed that Mor and the αCTD can interact with each other (3); and (iv) deletion analysis demonstrated that AT-rich sequences between −62 and −52 are important for full activity of P_m (2, 21). In this study, results from DNase I footprinting and gel mobility shift experiments demonstrate that the αCTD is the only polymerase component needed to generate the footprint just upstream from Mor and that Mor and the αCTD bind synergistically.

MATERIALS AND METHODS

Media, chemicals, and enzymes.

Luria-Bertani broth and agar (46) were used for standard bacterial growth; ampicillin at 50 μg/ml (US Biochemical Corp., Cleveland, Ohio) and chloramphenicol at 25 μg/ml (Sigma Chemical Co., St. Louis, Mo.) were added when necessary. Isopropyl-β-d-thiogalactopyranoside (IPTG) was obtained from US Biological. Radiolabeled compounds were from Perkin-Elmer Life Science (Boston, Mass.). In general, buffer components were obtained from Sigma Chemical Co., and gel reagents were purchased from Bio-Rad Laboratories (Hercules, Calif.). Phenol and phenol-chloroform were from Amresco Inc. (Solon, Ohio). DNase I (type II from bovine pancreas) was obtained from Sigma Chemical Co.; bovine serum albumin, phage T4 polynucleotide kinase (PNK), and deoxynucleoside triphosphates were purchased from Promega Corp. (Madison, Wis.). The enzymes Taq DNA polymerase, T4 DNA ligase, and shrimp alkaline phosphatase were from Roche (Basel, Switzerland). SeaKem ME and NuSieve GTG agarose were from FMC BioProducts (Rockland, Maine). The Ni-nitrilotriacetic acid column used for purifying the His-tagged α subunit of RNAP was from QIAGEN Inc. (Valencia, Calif.). Bench Mark protein size markers and low-DNA-mass ladder markers were obtained from Gibco-BRL Invitrogen (Carlsbad, Calif.). Oligonucleotides were purchased from Integrated DNA Technologies, Inc. (Coralville, Iowa).

Bacterial strains and plasmids.

Plasmid-containing derivatives of strain MH13312 [mcrA ΔproAB-lac thi gyrA endA hsdR relR supE44 recA/F′ (pro⁺ lacI^q1 ΔlacZY)] were used for isolation of plasmid DNA; MH13312 is a derivative of JM109 that contains an F′ pro⁺-lac element with lacI^q1 and ΔlacZY mutations (2). Strain RLG3538 [F⁻ ompT hsdS_B(r_B⁻m_B⁻) gal dcm λDE3/pLysS/pHTT7f1-NHα], obtained from W. Ross and R. Gourse, contains plasmid pHTT7f1-NHα in host strain BL21(DE3) (13), which expresses T7 RNA polymerase under the control of P_lacUV5; the plasmid contains the rpoA gene with an N-terminal His₆-tag under the control of the T7 gene 10 promoter and was used for overexpression and purification of the His-tagged α subunit of RNA polymerase (His-α) (50). Strain MH10713 [ompT hsdS_B(r_B⁻m_B⁻) gal dcm λDE3/pLysS], the Howe lab version of strain BL21(DE3) (31, 48), was transformed with plasmid pKM90 for Mor overproduction. Plasmid pKM90 contains the mor gene under the control of a T7 promoter in the pT7-7 expression vector and was used for overexpression and purification of the Mor activator protein (31).

Plasmid pMM1 was constructed by cloning P_m sequences from −98 to +10 into the EcoRI and BamHI sites of the plasmid vector pIA12; pIA12 is a HindIII linker-containing derivative of pLC1, which is a ΔlacY derivative of the promoter cloning vector pRS415 (2, 11, 44); pMM1 was used as the template when P_m probes were made by PCR.

Protein purification.

Overexpression and purification of His-α was performed essentially as described by Tang et al. (50). Strain RLG3538 was induced with IPTG, pelleted by centrifugation, resuspended in buffer A (50), and lysed by sonication, and the extract was clarified by centrifugation. Soluble His-α was purified by binding to Ni²⁺-nitrilotriacetic acid agarose beads, then washing with buffer A containing 20 mM imidazole, and eluting with buffer A containing 500 mM imidazole. After dialysis against storage buffer (25 mM Tris-HCl, pH 7.9, 100 mM NaCl, 0.1 mM EDTA, 0.1 mM dithiothreitol [DTT], 50% glycerol), the protein was aliquoted and stored at −20°C. The His-α concentration was determined by using a Bradford assay (8).

Mor protein was overexpressed in a strain made just before use by transformation of pKM90 into strain MH10713. Following sonication and centrifugation to remove cell debris, Mor protein was purified from the cell extract by sequential polyethyleneimine precipitation, ammonium sulfate precipitation (20%), and heparin agarose chromatography as described previously (2 and was generously provided by Y. Mo. The His-αCTD protein was overexpressed from a pET-28a (Novagen, EMD Biosciences) clone containing the sequence encoding amino acids 245 to 329 of the αCTD fused to a His₆ tag and thrombin cleavage site and located downstream from a T7 promoter. The His-αCTD protein was purified by Ni²⁺ affinity chromatography followed by size exclusion chromatography (Superdex 75); untagged αCTD was made by thrombin cleavage of the purified His-αCTD protein followed by Ni²⁺ affinity chromatography and size exclusion chromatography (Superdex 75). Purified His-αCTD and untagged αCTD were generously provided by M. Kumaraswami. Figure 1 shows a sodium dodecyl sulfate (SDS)-polyacrylamide gel (42) run with samples of each purified protein and stained with Coomassie brilliant blue dye.

FIG. 1.

An SDS-polyacrylamide gel containing samples of the purified proteins. Approximately 2.5 μg of each purified protein was subjected to electrophoresis on an SDS-polyacrylamide gel (12% acrylamide) and stained with Coomassie blue R250. Lanes were loaded with the following: protein markers whose masses (in kDa) are indicated at the left, lane 1; Mor, lane 2; His-αCTD, lane 3; αCTD, lane 4; and His-α, lane 5.

Probe preparation.

The 198-bp probe used for DNase I footprinting was generated by PCR using vector primers IRI21 (TGGGGATCGGAATTATCGT) and IRI22 (AACTGGCGGCTGTGGGATT), one of which was radiolabeled by T4 PNK using [γ-³²P]ATP (3,000 Ci/mmol), with plasmid pMM1 (containing P_m −98 to +10) DNA as template. After purification using a Qiaquick spin PCR purification kit (QIAGEN), probe concentration was estimated by comparing band intensities against a low-DNA-mass ladder run in parallel on a 2% agarose gel stained with ethidium bromide (42).

Short probes used in gel retardation assays were made by annealing pairs of oligonucleotides for 3 min at 55 or 60°C after preincubation for 5 min at 90°C. One strand of each pair was 5′-end-labeled with γ-³²P using T4 PNK and [γ-³²P]ATP (3,000 Ci/ml) prior to annealing. Probes were purified using a Qiaquick nucleotide removal kit (QIAGEN) before their concentrations were estimated by comparison to a low-DNA-mass ladder on a 4% agarose gel. Oligonucleotide pairs consisted of complementary sequences (top and bottom strands) containing the following P_m positions numbered relative to the transcription start site as +1: JM19-JM20, −54 to −21; JM21-JM22, −61 to −31; JM35-JM37, −57 to −31; and JM36-JM38, −59 to −31.

DNase I footprinting.

The binding reactions contained 1 to 2 ng of labeled probe and were performed in a total volume of 40 μl of binding buffer containing 20 mM Tris-HCl, pH 7.5, 50 mM NaCl, 5 mM MgCl₂, 0.1 mM EDTA, 2 mM CaCl₂, 1 mM DTT, 20 ng of bovine serum albumin/μl, 1 ng of calf thymus DNA/μl, and 7% glycerol. Various amounts of reconstituted RNAP (generously provided by W. Ross and R. Gourse), His-α, His-αCTD, untagged αCTD, or Mor were added to binding reactions in amounts based on binding efficiencies observed previously (2, 3, 7, 13, 50). Mor was always incubated with the probe for 3 to 10 min prior to addition of a second protein. After incubation, the reaction mixtures were treated with 4.5 ng of DNase I at room temperature for 20 to 45 s (depending on the proteins present), and reactions were stopped by addition of 50 μl of stop solution (3). The mixture was brought up to 200 μl with H₂O, extracted with phenol-chloroform, ethanol precipitated, washed with 70% ethanol, dried, and resuspended in loading buffer (42). Markers were generated by carrying out a Maxam-Gilbert G-only sequencing reaction with a similar probe (42). The products were separated by electrophoresis on a 6% sequencing gel (42) and visualized by autoradiography with Kodak BioMax film with or without an intensifying screen.

Gel mobility shift assays.

Gel shift assays were done using probes made by annealing primer pairs JM19-JM20, JM21-JM22, JM35-JM37, or JM36-JM38. Binding reactions were performed in 20 μl of buffer containing 20 mM Tris-HCl (pH 7.9), 50 mM NaCl, 1 mM DTT, 10% glycerol, and 0.7 to 2.0 nM labeled probe. Different amounts of Mor were added to achieve final concentrations of 68 nM and 136 nM, and the mixtures were incubated for 10 min at 30°C. Various amounts of purified His-α, His-αCTD, or αCTD were added, and incubation continued for another 10 min at 30°C. Mixtures were loaded onto 15-cm-long 6% acrylamide native gels containing 10% glycerol and 0.5× Tris-borate-EDTA buffer and subjected to electrophoresis in 0.5× Tris-borate-EDTA buffer containing 2% glycerol at 3°C for 2 to 3 h at 200 V (13). The gels were exposed to X-OMAT AR film (Kodak) without drying and with or without an intensifying screen.

Gel shift assays were also performed with the same 198-bp probes used for DNase I footprinting. In that case, 4% acrylamide gels were used, and electrophoresis was carried out for 1.5 h at 230 V.

RESULTS AND DISCUSSION

DNase I footprinting of P_m with RNA polymerase.

Previous DNase I footprinting with purified Mor and P_m DNA probe (3, 21) showed that Mor protected P_m positions −56 to −33 from digestion by DNase I (Fig. 2). The addition of RNA polymerase led to extension of the footprint downstream to at least +14 and the appearance of a new upstream footprint spanning positions −61 to −59 (3). The extended downstream footprint was probably due to open complex formation, since similar long footprints are typically found in open complexes (24, 39, 49), and melting of P_m DNA from −12 to −1 was detected by KMnO₄ footprinting of these complexes (4). It seemed most likely that the upstream footprint would result from binding of the αCTD of RNA polymerase to this region or from conformational changes in the DNA caused by open complex formation (3, 5, 41).

FIG. 2.

Middle promoter sequence and DNase I footprints. The sequence presented is that of the top strand from −66 to +23 of the Mu middle promoter P_m. The −10 hexamer is boxed, and the bent arrow represents the transcript initiating at +1. Dots mark 10-base intervals. The facing arrows identify the most important positions of the imperfect inverted-repeat sequences in the Mor binding site as defined by mutational analysis (2). The extent of the DNase I footprints observed due to binding of Mor alone or Mor and RNAP together are indicated by thick black bars. The sequences contained in short fragments used for gel mobility shift assays are shown by lines below the P_m sequence.

DNase I footprinting assays were used to distinguish between these possibilities. First, DNase I footprinting was performed at 30°C using a 198-bp DNA fragment containing P_m sequences −98 to +10, purified Mor protein, and two in vitro reconstituted RNA polymerases: wild-type RNAP and RNAP containing α subunits that were truncated at amino acid 235 and therefore deleted for the αCTD (RNAPΔαCTD) (Fig. 3). The footprint patterns for Mor alone and for Mor with wild-type RNAP at 30°C (Fig. 3B) were similar to those observed previously (3) and represent extended, heparin-resistant, open complexes melted from −12 through +4 (Y. Mo, personal communication). (Complete clearing of the downstream footprint was observed in other experiments [Y. Mo, personal communication]). Incubation of probe with RNAP alone caused only subtle changes in the DNase I band pattern (including slight hypersensitivity at −51 and −44 and slight protection at −36, −37, −23, and −24), indicating that any interactions between RNAP and P_m DNA in the absence of Mor were unstable. When footprinting was done with Mor and RNAP lacking the αCTD, neither the upstream nor the downstream footprints appeared (Fig. 3B), demonstrating that the αCTD was essential for stable binding of RNAP to P_m. Interestingly, the subtle changes in banding pattern observed with wild-type RNAP alone were also not observed with the deleted RNAP, suggesting that the αCTD may participate in the unstable RNAP-P_m DNA interactions as well.

FIG. 3.

DNase I footprinting of P_m with wild-type RNAP and RNAP lacking the αCTD. Before digestion by DNaseI linear 198-bp probe DNA, 5′-end-labeled on the top strand, was incubated for 10 min with 800 nM Mor; 30 (+), 60 (++), or 80 (+++) nM wild-type (WT) or mutant (ΔαCTD) RNAP reconstituted in vitro; or both Mor and RNAP. When both proteins were used, Mor was incubated with the probe for 3 min before RNAP was added; then incubation was continued for 10 min prior to DNase I treatment. The regions protected by Mor and RNAP are indicated by brackets, and the upstream footprint is indicated by a black bar. The numbers to the right of each panel correspond to positions in the P_m sequence, which are numbered relative to the transcription start site as +1. The presence or absence of protein is indicated by a + or −, respectively. Panel A also contains the products of a G-only Maxam-Gilbert sequencing reaction performed on a similar probe (differing only by the presence of additional Mu P_m sequence from +11 to +46) to serve as markers for determining the lengths of the DNase I fragments; the G-ladder fragments migrate 1.5 nucleotides faster than the corresponding DNase I fragments (45). Panels B and C contain DNase I digestion products generated by protein binding and DNase I digestion performed at 30 and 15°C, respectively.

To test whether the upstream footprint was caused by open complex formation, we repeated the DNase I footprinting at 15°C, since no open complex formation was observed in previous KMnO₄ footprinting experiments performed at 15°C (data not shown). The Mor footprint was the same as that observed at 30°C (Fig. 3C), and there was no extension of the footprint downstream when wild-type RNA polymerase was added, consistent with the absence of an open complex at 15°C. In contrast, the upstream footprint from −61 to −59 was still present, demonstrating that it is not open complex formation that generates the footprint. In other experiments performed similarly, the upstream footprints were sensitive to heparin (Y. Mo, personal communication), suggesting that they are in unstable closed complexes (24, 32). When RNAP lacking the αCTD was assayed, there was no upstream or downstream protection at all, again indicating that the αCTD is required for generation of both footprints.

DNase I footprinting of P_m with His-α, His-αCTD, and αCTD.

From the results of the above DNase I footprinting experiments, we concluded that the αCTD is essential for generation of the upstream footprint and that open complex formation is not required. To distinguish whether other RNAP subunits were involved or whether the αCTD was sufficient, we assayed for upstream protection from DNase I cleavage in experiments using purified His-α, His-αCTD, or untagged αCTD instead of RNA polymerase.

In the presence of Mor, all three αCTD-containing proteins generated an upstream footprint identical to that produced by the wild-type RNA polymerase at 15°C (Fig. 4). The footprint patterns observed with purified αCTD, His-αCTD, and His-α were indistinguishable, indicating that the αNTD was not required for αCTD binding and that the His tags did not interfere with binding. In the absence of Mor, none of the αCTD-containing proteins produced any detectable change in the DNase I band pattern. This experiment demonstrated that interactions between the αCTD, Mor, and DNA were both necessary and sufficient to generate the upstream footprint.

FIG. 4.

DNase I footprinting of P_m with His-α, His-αCTD, and αCTD. About 2 ng of the same probe used for Fig. 3B and C was incubated with 800 nM Mor, 9 μM His-α, 35 μM His-αCTD, or 50 μM αCTD for 10 min at 30°C. When Mor was used with His-α, His-αCTD, or αCTD, Mor was incubated with the probe for 5 min at 30°C prior to addition of the second protein, and incubation then continued for 10 min at 30°C prior to treatment with DNase I. The lane marked G contains the products of a G-only Maxam-Gilbert sequencing reaction performed on the same probe used for Fig. 3A. The P_m positions corresponding to the G-ladder bands are indicated on the right, and those for the DNase I digestion products are indicated on the left.

Gel mobility shift assay of His-α binding to P_m.

In DNase I footprinting experiments performed with less than saturating amounts of Mor protein, the Mor footprints were more clear in the presence of RNAP than in its absence (data not shown), leading us to think that Mor and RNAP might bind synergistically to P_m DNA, using protein-protein interactions to stabilize the binding of both proteins to the DNA. To test whether Mor and His-α might exhibit synergistic binding, we performed gel shift assays with the same labeled P_m DNA fragment used for DNase I footprinting in the presence of each protein, first alone and then together. Binding of Mor to the DNA probe resulted in generation of a typical slower-migrating binary complex, and the amount of complex increased with increasing Mor concentration (Fig. 5A). Binding of α is weak (7, 13), but a shifted species was observed for His-α alone when a high concentration was used. When both Mor and His-α were present, a supershifted species presumably corresponding to a His-α-Mor-DNA ternary complex was observed, even at low concentrations of Mor and His-α, concentrations that gave only a barely detectable (Mor) or undetectable (His-α) shift with each protein alone. When the assay was repeated with Mor and the αCTD instead of His-α, similar results were observed (Fig. 5B). In other experiments involving incubation of His-Mor and His-αCTD with short P_m DNA fragments, the complexes formed were purified and shown by SDS-polyacrylamide gel electrophoresis to contain both His-Mor and His-αCTD (M. Kumaraswami, personal communication). Therefore, taken together with the results of previous yeast interaction trap assays (3), these experiments demonstrate that Mor and α do bind synergistically to P_m, forming a stable ternary complex, and that at least part of that synergistic binding is due to interactions between Mor and the αCTD.

FIG. 5.

Gel mobility shift assays of Mor, His-α, and αCTD binding to P_m DNA. (A) The same 198-bp probe used for DNase I footprinting was incubated with Mor at 68 (+) or 200 (++) nM or 1 μM (four +'s) or His-α at 2.7 (+), 5.4 (++), or 10.8 (four +'s) μM, either alone or together. When alone, binding reactions were incubated for 10 min at 30°C; when together, Mor was preincubated with probe for 10 min at 30°C, then His-α was added, and incubation was continued for 10 min at 30°C prior to loading on the gel. (B) The assay was performed as described for panel A except purified αCTD was used at a concentration of 30 μM (+) in place of His-α. A − indicates that no protein was added. Bands corresponding to free probe, the Mor-DNA or His-α-DNA binary complexes, and the Mor-His-α-DNA or Mor-αCTD-DNA ternary complexes are indicated by the labels F, B, and T, respectively, to the left of each panel.

Examination of the P_m sequence (Fig. 2) reveals the presence of a second AT-rich region, the stretch of seven T residues at positions −30 through −24, which might serve as an additional binding site for the αCTD. In the DNase I footprinting experiments, this region had no bands in the DNA-only control lane, making it impossible to detect a footprint if the αCTD bound there. To test for possible αCTD binding to that region, we performed gel shift assays with short DNA fragments made by annealing pairs of complementary oligonucleotides. One 31-bp fragment, with P_m sequence −61 to −31, contained the Mor binding site and upstream αCTD binding site. The second probe, a 34-bp fragment, with P_m sequence −54 to −21, contained the Mor binding site and downstream T tract but lacked the upstream αCTD binding site. Mor bound normally to both probes (Fig. 6A), but only the −61 to −31 probe containing the upstream binding site gave a supershifted species reflecting binding of both Mor and His-α. The absence of a supershifted species with the −54 to −21 probe indicates either that the T tract in P_m does not serve as an efficient binding site for α or that this site is too distant and on a different face of the helix from the Mor binding site, preventing the Mor and His-α synergistic binding needed for stability of the ternary complex.

FIG. 6.

Gel mobility shift assays with short P_m DNA fragments. Double-stranded blunt-ended P_m DNA probes made by annealing of complementary oligonucleotides were incubated with Mor at 68 (+) or 136 (++) nM or 0.5 (+), 1 (++) or 2.7 (four +'s) μM His-α in 20 μl for 10 min at 30°C. For lanes containing both proteins, 68 nM Mor was incubated with probe for 10 min at 30°C prior to addition of His-α, and then the mixture was incubated for another 10 min at 30°C before loading on the gel. The notations are the same as those in Fig. 5 except that the P_m sequence contained in each probe is indicated above the set of lanes containing that probe. Panels A and B were done separately with probes of different lengths; the bands below the free probe are probably excess unannealed oligonucleotide.

The −61 to −31 probe used in the experiment described above contained 10 bp beyond the end of the Mor dyad-symmetry element (Fig. 2). To determine how much of this region is needed for αCTD binding, we repeated the gel shift assays with shorter probes containing P_m sequences −59 to −31 and −57 to −31. The longer −59 to −31 probe gave strong supershifted bands comparable to those seen with the −61 to −31 probe, whereas the −57 to −31 probe gave a weaker supershifted species whose migration was retarded much less than that of the others (Fig. 6B). We conclude from this experiment that 7 to 8 bp beyond the Mor dyad-symmetry element are required for efficient ternary complex formation and stability. This length is similar to that of the proximal subsite of the bipartite UP element (12). The short length of the upstream footprint is also consistent with the conclusion that only a single αCTD is bound and that αCTD binding occurs just upstream from Mor.

In conclusion, the results of DNase I footprinting and gel mobility shift assays demonstrate that the αCTD of RNAP binds just upstream of the Mor binding site in P_m, exhibiting synergistic binding with Mor. Given the requirement for the αCTD for normal P_m activation (3) we predict that these synergistic binding interactions play an important role in the activation process.