Localization of a ς70 Binding Site on the N Terminus of the Escherichia coli RNA Polymerase β′ Subunit*

The Escherichia coli genome encodes genes for seven different ς subunit species while only having single genes for the α, β, and β′ subunits that make up the RNA polymerase core enzyme. The various ς factors compete for binding to the core enzyme, upon which they confer promoter DNA-specific transcription initiation to the polymerase. We have mapped a major interaction site between one of the ς species, ς70, and β′. Using far-Western blotting analysis of chemically cleaved and genetically engineered protein fragments, we have identified a N-terminal fragment of β′ (residues 60–309) that could bind ς70. We were able to more precisely map the interaction domain to amino acid residues 260–309 of β′ using nickel nitrilotriacetic acid co-immobilization assays.

The RNA polymerase of Escherichia coli is a large, multisubunit enzyme existing in two forms. The core enzyme, consisting of subunits ␤ and ␤Ј and an ␣ subunit dimer, carries out processive transcription elongation followed by termination. When one of a variety of sigma factors is added to core, the holoenzyme is formed. The subunit confers promoter-specific DNA binding and transcription initiation capabilities to the enzyme (1)(2)(3). 70 of E. coli was the first factor to be described and characterized (4). Since then, numerous 's have been discovered throughout the Eubacterial kingdom, including six alternative Јs in E. coli. Each directs transcription initiation from a specific set of promoters to transcribe genes usually with related functions. This control of transcription is mediated partially through the competition of the individual factors for the core enzyme and is a major part of global gene regulation in bacteria (5). Elucidation of the structural characteristics of the core RNA polymerase-factor binding interaction will be very beneficial in fully understanding this aspect of regulation.
As the number of identified factors increased, it became apparent that they shared several regions of amino acid sequence similarity (1,6,7). Work has been ongoing to assign specific functions to these conserved regions (8 -12). Deletion analysis of 70 identified a segment of the protein that overlaps region 2.1 (residues 361-390) as being necessary and sufficient for core binding (12). A mutation in a homologous region of Bacillus subtilis E has also been shown to affect core binding (13). However, recent findings of core binding mutations in other conserved and nonconserved regions of 32 have led to the idea of multiple binding sites for the subunit on the core enzyme (14 -16).
To date, little is known about the location of the binding sites on the core subunits. There have been two observations that have identified deletions in the ␤ or ␤Ј subunits that produce subunits still capable of forming core enzyme structures but not holo. First, a ␤ subunit truncation, missing approximately 200 amino acids (aa) 1 of the C terminus, was shown by glycerol gradient centrifugation to migrate with the other core subunits but was never seen in the -containing fractions (17). Second, when immunoprecipitation assays were performed using reconstituted RNA polymerase containing ␤Ј deletion mutants missing aa 201-477, the core subunits were recovered in the same fraction but lacked (18). This idea that binding is affected by perturbations of the C terminus of ␤ and the N terminus of ␤Ј is consistent with experiments showing that these two subunit termini are physically close together and can be fused through a flexible linker and still form a functional enzyme (19). Recent protein-protein footprinting data have identified a similar region on ␤Ј and two new sites on ␤ for possible interactions with the 70 subunit (20).
The ␤ and ␤Ј subunits each contain regions that have high sequence homology with the two largest subunits of eukaryal polymerases (21)(22)(23). Some of these conserved regions may act as interaction domains. We define an interaction domain as the minimal region of a protein that can independently fold to form the secondary and tertiary structure required to interact with another protein, DNA, RNA, or ligand. Interaction domains will always be larger than the actual binding site, the amino acids in direct contact with the binding partner. Therefore, additional work will be needed to identify the critical residues involved in making binding contacts. Severinov et al. (24 -26) demonstrated the domain-like properties of ␤ and ␤Ј by reconstitution of functional RNA polymerase from fragmented ␤ and ␤Ј subunits. This indicates that the properties of the protein do not require the entire intact length of the subunit but rather can be generated with smaller domain modules.
In this study, we set out to map the protein-protein interaction domains on both ␤ and ␤Ј required for the binding of 70 . Chemical and enzymatic cleavage as well as PCR methods were used to generate fragments of ␤ and ␤Ј for use in mapping the interaction domains. Using far-Western blotting and nickel nitrilotriacetic acid (Ni 2ϩ -NTA) co-immobilization assays (27)(28)(29), we were able to map a strong specific binding site for 70 to the N terminus of ␤Ј. This report shows that this binding site is located within a span of residues (260 -309) that overlaps conserved region B of ␤Ј (23).

Construction of Plasmids
Plasmid characteristics are described in Table I. An overexpression vector for C-terminal hexahistidine (His 6 )-tagged ␤Ј (pTA500) was constructed by removing the XbaI-HindIII fragment from pRL663 and placing it in pET28b (Novagen) (31). N-terminally His 6 -tagged ␤Ј was expressed from pTA499 that was constructed using PCR to place the His 6 tag on the N terminus of a fragment that overlapped the NruI site of ␤Ј. This fragment was placed into the pET28b vector followed by the insertion of the C-terminal portion of the gene on a NruI-HindIII fragment from pRL663. The C-terminal His 6 tag from the pRL663 fragment was removed by replacement of the RsrII-HindIII fragment with a PCR product coding for the wild type C terminus. pTA501 was constructed by creating an N-terminal His 6 tag via PCR for a fragment of ␤, which overlapped the KpnI site of ␤. The fragment was placed into the pET28b vector. The C terminus of the gene was added by insertion of a KpnI-HindIII fragment containing the wild type coding sequence. pTA502, coding for the C-terminal His 6 -tagged ␤ subunit, was derived using PCR to insert a N-terminal NcoI site onto a fragment overlapping KpnI. The C-terminal His 6 -containing fragment was inserted on a KpnI-HindIII fragment from pRL706 (19). Vectors expressing unmodified fragments of ␤Ј were obtained by PCR cloning of the desired fragment and placement of the fragment into either pET21a (Novagen) for pTA528, pTA530, pTA535, and pTA536 or pET24a (Novagen) for pTA519 using NdeI and XhoI restriction sites. pTA522-525, pTA531, and pTA533 were all created by amplifying the specified ␤Ј region via PCR and inserted into a pET21a derivative that had been modified to fuse a N-terminal His 6 and heart muscle kinase (HMK) recognition site to the expressed proteins. pTA532 and pTA534 were constructed in the same fashion with the exception that the His 6 -HMK vector derivative was constructed from pET28b. pTA547-549 were created by inserting the fragments, N-terminally truncated via PCR, that overlapped the SnaBI site of ␤Ј into pET24a. The C-terminal coding region of the gene was inserted on a SnaBI-HindIII fragment from pTA500. pTA546 was created by fusing a C-terminal His 6 tag directly after residue 309 via PCR. The fragment was placed into the pET24a vector using NdeI and XhoI sites. To use 70 as a radioactive probe, the HMK site was fused to the N terminus of 70 along with a His 6 purification tag. pHMK-His 6 -70 was created by placing the 70 gene into a derivative of pET28b vector that contained the N-terminal His 6 and HMK fusion and adds a total of 13 extra aa (MHHHHHHARRASV) to the N terminus of 70 . All products created by PCR were sequenced to ensure that no mutations had been introduced.

Expression and Purification of Proteins
Plasmids were transformed into BL21(DE3) (Novagen) for expression (31). The cells were grown in 1-liter cultures at 37°C in LB medium with either 100 g/ml ampicillin or 50 g/ml kanamycin. The cultures were grown to an A 600 between 0.6 and 0.8 and then induced with 1 mM isopropyl-␤-D-thiogalactopyranoside. Three hours after induction, the cells were harvested by centrifugation at 8,000 ϫ g for 15 min and frozen at Ϫ20°C until use.
The cells were thawed and resuspended in 10 ml of lysis buffer (40 mM Tris-HCl, pH 7.9, 0.3 M KCl, 10 mM EDTA, and 0.1 mM phenylmethylsulfonyl fluoride), and lysozyme was added to 100 g/ml. The cells were incubated on ice for 15 min then sonicated three times in 60-s bursts. The recombinant protein in the form of inclusion bodies was separated from the soluble lysate by centrifugation at 27,000 ϫ g for 15 min. The inclusion body pellet was resuspended, by sonication, in 10 ml of lysis buffer ϩ 2% (w/v) sodium deoxycholate. The mixture was centrifuged at 27,000 ϫ g for 15 min and the supernatant discarded. The deoxycholate-washed inclusion bodies were resuspended in 10 ml deionized water and centrifuged at 27,000 ϫ g for 15 min. The water wash was repeated, and the inclusion bodies were aliquoted into 1-mg pellets and frozen at Ϫ20°C until use. 70 inclusion bodies were solubilized, refolded, and purified according to a variation of the procedure of Gribskov and Burgess (32). The inclusion bodies were solubilized by resuspension in 6 M guanidine HCl. The proteins were allowed to refold by diluting the denaturant 64-fold with buffer A (50 mM Tris-HCl, 0.5 mM EDTA, and 5% (v/v) glycerol) in 2-fold steps over 2 h. One gram of DE52 resin (Whatman) was added and mixed with slow stirring for 24 h at 4°C. The resin was then collected in a 10-ml column, washed, and the protein eluted with a gradient from 0.1 to 1 M NaCl in buffer A. The 70 fractions were pooled and dialyzed overnight against 1 liter of storage buffer (50 mM Tris-HCl, 0.5 mM EDTA, 0.1 M NaCl, 0.1 mM DTT, and 50% (v/v) glycerol) and stored at Ϫ20°C. Whole cell lystates were prepared as follows. Cells containing truncated ␤Ј expression plasmids were grown to an A 600 of 0.6 -0.8 and induced with 1 mM isopropyl-␤-D-thiogalactopyranoside. The cells were grown for an additional 30 min. A 200-l sample was removed and sonicated 3 ϫ 30 s. 20 l of glycerol and 20 l of SDS-sample buffer were added and heated for 2 min at 95°C then stored at Ϫ20°C until use.

Protein Cleavage
␤ and ␤Ј inclusion bodies were subjected to chemical or enzymatic cleavage (see below) and then purified by nickel affinity chromatography as follows. The cleavage reaction was loaded onto 300 l of Ni 2ϩ -NTA resin (Qiagen) in a Bio-Rad mini-column. The resin had been pre-equilibrated with buffer B (20 mM Tris-HCl, pH 7.9, 500 mM NaCl, 5 mM imidazole, 0.1% (v/v) Tween 20, and 10% (v/v) glycerol) ϩ 8 M urea. The protein bound resin was washed with 10 column volumes of buffer B ϩ 8 M urea followed by 10 column volumes of buffer B to allow refolding. The resin was then washed with 500 l of buffer B ϩ 40 mM imidazole. The protein was eluted with 500 l of buffer B ϩ 200 mM imidazole. The eluted fractions were stored at Ϫ20°C.
NTCB Cleavage (33)-1 mg of inclusion body protein was resuspended in 1 ml of buffer B ϩ 8 M urea. DTT was added to 5-fold molar excess over the thiol groups in the protein. The mixture was incubated for 15 min at 37°C to reduce any disulfide bonds. NTCB was added to 5-fold molar excess over total sulfhydryl groups. The pH was adjusted to 9.5 with NaOH. The reaction mixture was incubated for 2 h at room temperature. The cleavage mixture was diluted 1:10 in buffer B ϩ 8 M urea and loaded onto a Ni 2ϩ -NTA column as described above.
Hydroxylamine Cleavage (34)-1 mg of inclusion body protein was resuspended in 1 ml of buffer B plus 8 M urea. Five-hundred microliters of the solubilized protein were added to 500 l of hydroxylamine cleavage solution (0.4 M CHES, pH 9.5, 4 M hydroxylamine HCl) and incubated 2 h at 42°C. ␤-Mercaptoethanol was added to 0.1 M and incubated 10 min at 37°C. The mixture was diluted 1:10 in buffer B ϩ 8 M urea and loaded onto a Ni 2ϩ -NTA column as described above.
Thermolysin Cleavage (35)-1 mg of inclusion body protein was resuspended in 100 l of buffer B ϩ 8 M urea and incubated for 15 min at 37°C. Thermolysin was added at protein:protease ratios of 4,000:1, 8,000:1, and 16,000:1 (w/w). Reactions were carried out for 30 min at room temperature. The reactions were loaded onto a Ni 2ϩ -NTA column as described above.
Trypsin Cleavage (35)-1 mg of inclusion body protein was resuspended in 1 ml of buffer B ϩ 8 M urea and incubated for 15 min at 37°C. The mixture was diluted to 4 M urea by adding an equal volume of buffer B. Trypsin was added at protein:protease ratios of 4,000:1, 8,000:1, and 16,000:1 (w/w). Digestion was performed at room temperature for 30 min. The reactions were loaded onto a Ni 2ϩ -NTA column as described above.
Gel Blot-Protein cleavage fragments or whole cell lysates were separated by SDS-polyacrylamide gel electrophoresis (PAGE). The proteins were electrophoretically transferred onto 0.05-m nitrocellulose. The nitrocellulose was blocked by incubating in HYB buffer for 16 h at 4°C.
Labeling-Labeling of 70 was done in a 100-l reaction volume. 50 l of 2 ϫ kinase buffer (40 mM Tris-HCl, pH 7.4, 200 mM NaCl, 24 mM MgCl 2 , 2 mM DTT, and 50% (v/v) glycerol) was added to 50 g of 70 protein. 240 units of cAMP-dependent kinase-catalytic subunit (Promega) was added, and the total volume was brought up to 99 l with deionized water. One microliter of [␥-32 P]ATP (0.15 mCi/l) was added. The mixture was incubated at room temperature for 30 min. The reaction mixture was then loaded onto a Biospin-P6 column (Bio-Rad) pre-equilibrated with 1 ϫ kinase buffer and spun at 1,100 ϫ g for 4 min. The flow-through was collected and stored at Ϫ20°C.
Probing-The blocked nitrocellulose was incubated in 10 ml of HYB buffer with 4 ϫ 10 5 cpm/ml 32 P-labeled 70 for 3 h at room temperature. The blot was washed three times with 10 ml of HYB buffer for 3 min each. The blot was then dried and exposed to film or PhosphorImager (Molecular Dynamics).

Co-immobilization
One milligram of His 6 -tagged, truncated ␤Ј was solubilized in 1 ml of buffer C (20 mM Tris-HCl, pH 7.9, 200 mM NaCl, 5 mM imidazole, 0.1% (v/v) Tween 20, and 10% (v/v) glycerol) ϩ 8 M urea. Twenty micrograms of the protein solution were loaded onto 150 l of Ni 2ϩ -NTA resin. The column was washed with 15 column volumes of buffer C ϩ 8 M urea, followed by a 15-column volume wash with buffer C to allow refolding. Then 30 g of native 70 were loaded onto the column. The column was washed with 20 volumes of buffer C. The bound proteins were eluted with 300 l of buffer C ϩ 250 mM imidazole. Samples from the 70 flow-through, wash, and elution fractions were analyzed by SDS-PAGE.

RESULTS
70 Interacts Strongly with ␤Ј Subunit and Weakly with ␤ Subunit in Far-Western Blot Analysis-Previously we developed a method that could be used to map antibody epitopes using histidine-tagged purification of partially cleaved proteins (35). To apply this method for mapping the protein-protein interactions of 70 on ␤ or ␤Ј, we needed to determine whether 70 could bind to individual ␤ and ␤Ј subunits outside of the core enzyme complex. Far-Western assays of dot blots were used to assess the binding. Inclusion body proteins of ␤ and ␤Ј were separately solubilized in urea and spotted onto nitrocellulose. Bovine serum albumin (BSA) was spotted as a control for nonspecific binding. The nitrocellulose was blocked and the denaturant washed away. The blot was then probed with 32 Plabeled 70 . Both ␤ and ␤Ј subunits bound 70 , the BSA control did not (Fig. 1). Identical dot blots were probed with control solutions lacking either the kinase or 70 to ensure that the signal was not due to nucleotide binding or phosphorylation of the ␤ or ␤Ј subunits. Neither control blot produced a detectable signal (data not shown). Thus, both ␤ and ␤Ј subunits can individually bind 70 . 70 Interaction Specific for ␤/␤Ј in Far-Western Analysis-We performed an additional test to assess the specificity of the far-Western analysis using 70 as a probe. A cell lysate from a log phase culture was separated by SDS-PAGE, blotted onto nitrocellulose, and probed with 70 . The only strong signal produced had the same mobility as ␤ and ␤Ј (Fig. 2B, lanes C  and L). The absence of other strong signals indicates that 70 is not binding nonspecifically to ␤ and/or ␤Ј. Minor bands were observed as expected, since there are other proteins that have been shown to interact with 70 (activators, anti-, etc.) (36,37).
A Strong, Specific Binding Site for 70 Is Located in the N Terminus of the ␤Ј Subunit-To map the 70 interaction sites on ␤ and ␤Ј, we performed far-Western analysis of chemical cleavage products of the two large subunits (Fig. 3). The amino acid sequences of both were analyzed using MacVector software (Oxford Molecular Group) to identify specific chemical cleavage sites. Based on this analysis, cleavage reagents were chosen that produced an array of products following partial digestion that provide the highest resolution for mapping. Both N-and C-terminal His 6 -tagged constructs of ␤ and ␤Ј were subjected to cleavage under denaturing conditions. The products of the cleavage reaction were purified under denaturing conditions using Ni 2ϩ -NTA resin to isolate cleavage fragments containing a His 6 tag. These purified fragments were then identified based on their mobility in SDS-PAGE, and their exact size was determined based on the cleavage site which produced them (Fig. 3A). When the cleavage fragments were fractionated by SDS-PAGE, they produced a ladder of descending sized fragments with a common end (either N or C terminus depending on the placement of the His 6 tag). The use of both Nand C-terminally His 6 -tagged fragments allows the positive identification of both the N and C termini of the interaction domain. The 70 probe will only bind the fragments that have an intact interaction domain. The N-terminally His 6 -tagged ␤Ј ladders produced by hydroxylamine and NTCB cleavage both contained several fragments that retained the ability to bind 70 (Fig. 3B, lanes 3 and 7). Thus, a large portion of the C terminus of ␤Ј can be removed without affecting 70 binding. The smallest fragment to bind 70 was the 1-309 aa fragment of ␤Ј in the hydroxylamine ladder (Fig. 3B, lane 3). In the C-terminally His 6 -tagged ladders, only full-length ␤Ј bound 70 (Fig. 3B, lanes 4 and 8). These results indicated that a strong specific binding site is located within amino acids 1-309 of ␤Ј (␤Ј 1-309 ). The ␤ fragment ladders failed to produce signals strong enough to effectively map the interaction domain. For the remainder of this study, we will focus only on mapping the 70 -␤Ј interaction domain. The resolution of chemical cleavage mapping was relatively low due to the limited number of cleavage sites on ␤Ј for the reagents available. To increase the number of proteolytic fragments that could be used in mapping, we used enzymatic cleavage. Specificity of cleavage by many proteases is not as limited as with chemical cleavage reagents. Therefore, there are many more sites of cleavage and more fragments are produced. Partial digests of N-and C-terminally His 6 -tagged ␤Ј were conducted using trypsin and thermolysin. The fragments were again purified, blotted, and probed with 70 . However, even with the increased number of fragments, the interaction domain could not be narrowed from its previous length of 1-309 aa (data not shown).
Interaction Domain Narrowed to 60 -309 aa by Far-Western Blotting with Truncated Fragments-In trying to define this binding site more precisely, we made various truncated fragments using PCR. Using the ␤Ј 1-309 fragment as a starting point, we made constructs that were truncated at either the N or C terminus. DNA coding for the truncated fragments was cloned into overexpression plasmids. When cells containing these plasmids had been grown to an A 600 of 0.6, expression was induced. The cells were only allowed to grow for 30 min after induction. A whole cell lysate from each culture was made and used for far-Western blotting assays (Fig. 4). Short expression times kept the expression level of the induced protein comparable with the other proteins in the lysate. The use of the whole cell lysate in far-Western blotting assays was an internal control to ensure binding was specific for the protein of interest. This also meant that the various proteins would not have to be purified and could be expressed without purification tags. When constructs were made where the C terminus of ␤Ј 1-309 was truncated beyond amino acid 300, the binding of 70 was lost (Fig. 4A). However, the N terminus of the same fragment could be truncated up to 60 aa without diminishing the signal (Fig. 4B). ␤Ј 100 -309 still showed binding, but at a lower level, and ␤Ј 150 -309 did not bind 70 . These results narrowed the 70 binding site to ␤Ј 60 -309 . Western blot experiments using anti-␤Ј monoclonal antibodies were done to ensure that the protein fragments were being transferred to the nitrocellulose and that they were fragments of ␤Ј (data not shown).
Co-immobilization Assays Further Narrow Interaction Site to 260 -309 aa of ␤Ј-Ni 2ϩ -NTA co-immobilization assays were used to confirm and extend the results that had been produced using far-Western blotting. The proteins to be assayed for binding 70 were fused to His 6 purification tags and overexpressed in the form of inclusion bodies. The inclusion body protein was solubilized with 8 M urea and loaded onto Ni 2ϩ -NTA resin. The denaturant was washed away allowing the proteins to refold while still remaining bound to the resin. Native 70 was then loaded onto the column. The column was washed, and the bound proteins were then eluted with imidazole. Any truncated protein that contained the interaction domain for 70 would cause 70 to be bound and to be in the eluted fraction. The results of the these binding experiments are consistent with the far-Western blotting experiments in respect to defining the C-terminal boundary of the domain. ␤Ј 1-309 bound 70 , while ␤Ј 1-300 and ␤Ј 1-280 did not bind 70 (Fig. 5A-C). Refolded ␤Ј 1-309 without a His 6 tag was mixed with 70 and passed over the Ni 2ϩ -NTA to ensure the complex was not nonspecifically binding to the column. The complex passed through the column and was not seen in the eluted fraction (Fig. 5D). As a control, BSA was loaded onto a column containing ␤Ј 1-309 (Fig. 5E). BSA was seen only in the flow-through and not in the eluted fraction, suggesting ␤Ј 1-309 binds 70 specifically.
For the N-terminal boundary, the results showed that more of the N terminus could be removed without affecting 70 binding than was seen by the far-Western assay. Several Nterminally truncated fragments, all having aa 309 as the Cterminal boundary followed by a His 6 tag, were constructed and used in co-immobilization assays. Truncations to residues 33, 60, 100, 178, and 200 still produced fragments capable of binding 70 (data not shown). ␤Ј 260 -309 , the smallest fragment we could make and still manipulate efficiently in our assay, retained the ability to bind 70 (Fig. 6A). To find the N terminus of the interaction domain, truncations greater than residue 240 were made from full-length ␤Ј. A truncation of the first 260 residues of ␤Ј (␤Ј 260-C ) bound 70 , while ␤Ј 270-C showed diminished binding, and ␤Ј 280-C showed no detectable binding of 70 (Fig. 6, B-D). Taken together these results indicate that a strong 70 binding site on the core polymerase is located within the aa 260 -309 region of ␤Ј.

DISCUSSION
Since the discovery of the subunit in 1968, a great deal of effort has gone into characterizing the interactions between and the core enzyme (4). To date, several biochemical and genetic studies have contributed to what is known about the putative core binding domains on , however, much less is known about the sites on core that bind (2). In the holoenzyme assembly pathway, ␤Ј is added to the ␣ 2 ␤ complex and then is added to form the holoenzyme (38). This would suggest that either the major binding site is located on ␤Ј or is formed in cooperation with ␣ and/or ␤ upon ␤Ј assembly into the core enzyme. The isolation of 70 ⅐␤Ј complexes provides evidence for the former (18). In this report we have localized a strong binding site for 70 on ␤Ј, as well as identified low level binding affinity for 70 to ␤. Thus, we conclude that ␤Ј provides the major binding interaction for 70 in the holoenzyme while ␤ adds a secondary binding interaction. Multiple core binding sites on have been suggested in light of mutations apparently affecting core binding that map outside of conserved region 2.1 (14 -16).
The primary finding of this work is that a strong binding site for 70 is located within residues 260 -309 of ␤Ј. A deletion of residues 201-477 of ␤Ј has been reported previously to produce a mutant protein that could still form core but not holoenzyme (18). The problem with such deletion studies is that one cannot conclude that the binding site is located in the region deleted, but merely that the region, when deleted, prevents correct formation of the interaction domain. Results obtained from protein-protein footprinting experiments indicated that a similar region of ␤Ј (residues 228 -461) was physically close to 70 (20). There is difficulty in interpreting these results, since the assay gives indications of physical proximity of the proteins that do not necessarily correspond to protein-protein binding. From our findings it can be concluded that a major 70 binding site is located within these regions.
The 70 interaction domain on ␤Ј contains several residues located in conserved region B (23). This region does not have any known function. Secondary structural predictions derived from the PHD program (39) for residues 260 -309 indicates one helix from residue 264 to residue 283 connected by a loop to a second helix from residue 292 to residue 309. These predicted helices are also predicted to form coiled coils (40). This is of particular interest, since similar predictions were made for residues 355-391 of 70 . These residues overlap conserved region 2.1. The crystal structure of the protease-resistant fragment of 70 confirmed the prediction that the helix containing region 2.1 is forming a coiled coil with conserved region 1.2 (41). Since coiled coils have been shown to be involved in many protein-protein interactions (42)(43)(44), this would suggest that ␤Ј 260 -309 may be interacting with region 2.1 of 70 . Our laboratory is investigating where the ␤Ј 260 -309 interaction domain is contacting 70 .
Previously in our laboratory we have developed a method for quickly mapping epitopes for monoclonal antibodies (35). We describe here an application of that method to identify domains involved in other protein-protein interactions. Ordered fragment ladder far-Western blotting was used to map the 70 binding site on ␤Ј to within ␤Ј 60 -309 . This method relies on the fact that after the removal of the denaturant some fraction of FIG. 3. Ordered fragment far-Western. ␤ and ␤Ј subunits were cleaved with hydroxylamine and NTCB. The Ni 2ϩ -NTA purified fragments were separated on identically loaded 8 -16% SDS-PAGE gels (Novex). One gel (A) was stained with Coomassie Blue, and the other (B) was transferred to nitrocellulose and probed with 32 P-70 . Lanes: 1, markers; 2, E. coli lysate; 3, N-His 6 ␤Ј-hydroxylamine; 4, C-His 6 ␤Ј-hydroxylamine; 5, N-His 6 ␤-hydroxylamine; 6, C-His 6 ␤-hydroxylamine; 7, N-His 6 ␤Ј-NTCB; 8, C-His 6 ␤Ј-NTCB; 9, N-His 6 ␤-NTCB; 10, C-His 6 ␤-NTCB; 11, purified core polymerase. the blotted protein will be able to refold and produce the proper conformation for binding of the probe. The specificity of the assay was demonstrated by probing whole cell lysates and identifying ␤Ј as the major binding interaction. The combination of specific chemical cleavage of proteins and far-Western blotting provided a very rapid and effective way to localize this protein-protein interaction. Cloning and screening individual, truncated fragments was necessary only after the interaction domain had been targeted. Having to make truncations of ␤Ј all along its length would have been a long and tedious process. The protein cleavage and Ni 2ϩ column purification procedures can be done in one day, thus making the assay more expedient and less tedious.
To confirm and extend the results obtained with far-Western blotting, Ni 2ϩ co-immobilization assays were performed. These experiments also demonstrated that fragments from the N terminus to residue 309 could still bind 70 , while removal of just 9 C-terminal residues to aa 300 would abolish binding. The results obtained from the N-terminally truncated fragments in these assays gave much better resolution of the binding site location than was obtained from far-Western assays. Up to 260 residues could be removed from the N terminus without affecting 70 binding. When 270 residues were removed, binding of 70 was diminished but not abolished, suggesting that either part of the binding site had been removed or the binding site was intact but hindered from refolding due to the loss of upstream residues. To ensure that the binding site was what we were actually mapping and not just a region required for proper folding of the actual binding site, protein fragments were made from 260 -309 of ␤Ј and shown to be sufficient for binding. We believe that the difference in the identified interaction domain size between the far-Western assay (␤Ј 60 -309 ) and the co-immobilization assay (␤Ј 260 -309 ) is consistent with the properties of each assay. The far-Western assay requires the interaction domain to refold and properly present the binding site while some portion of the protein is attached to the nitrocellulose membrane. We believe that as such the proteins are more conformationally restricted than proteins bound only at one terminus as in the Ni 2ϩ -NTA co-immobilization assay. Therefore, more of the protein length is required to form a scaffoldlike structure to keep the interaction domain away from the membrane surface. The combination of mapping methods provides a rapid, high resolution procedure for identification of protein interaction domains.
We have work in progress to map the binding sites on core for the other 6 known E. coli factors. It has been hypothesized that the core binding region on the factors must be highly conserved, since they all must bind core polymerase (1,2,7). This has led to region 2 as the primary candidate, since it is the most highly conserved region. This suggests that all of the factors in the cell may bind to the same site or sites on core polymerase. We are now in a position to test this hypothesis and identify if the other factors can also bind ␤Ј 260 -309 .
Acknowledgments-We thank Robert Landick for generously providing plasmids pRL663 and pRL706. We also thank B. Pietz, V. Svetlov, and N. Thompson for technical advice and critical reading of this manuscript.

FIG. 5.
Co-immobilization with C-terminally truncated fragments. A-C, fragments of ␤Ј were bound to a Ni 2ϩ -NTA column, and 70 was passed over the column. The flow-through (ft) was collected. The columns were washed (w) with 20 column volumes then eluted (e) with 250 mM imidazole. The fractions were run by SDS-PAGE using 8 -16% Tris-glycine gels. Only for ␤Ј fragments capable of binding 70 will 70 be seen in the elution fraction. D, as a control for nonspecific binding of either 70 or the 70 -␤Ј 1-309 complex, a non-His 6 -tagged ␤Ј 1-309 was added to 70 and passed over a Ni 2ϩ -NTA column. E, BSA was passed over a Ni 2ϩ -NTA column that had ␤Ј 1-309 bound as another control for nonspecific binding.

FIG. 6.
Co-immobilization with N-terminally truncated fragments. Fragments of ␤Ј were bound to a Ni 2ϩ -NTA column, and 70 was passed over the column. The flow-through (ft) was collected. The columns were washed (w) with 20 column volumes and then eluted (e) with 250 mM imidazole. Only for ␤Ј fragments capable of binding 70 will 70 be seen in the elution fraction. A, the ␤Ј 260 -309 fragment co-immobilization assay fractions separated on a 10 -20% Tris-Tricine polyacrylamide gel (Novex). B-D, specified co-immobilization fractions separated on and 10 -20% Tris-glycine polyacrylamide gel (Novex).