Disruption of Escherichia coli HepA, an RNA Polymerase-associated Protein, Causes UV Sensitivity*

During the development of purification procedures for Escherichia coli RNA polymerase (RNAP), we noticed the consistent co-purification of a 110-kDa polypeptide. Here, we report the identification of the 110-kDa protein as the product of thehepA gene, a member of the SNF2 family of putative helicases. We have cloned the hepA gene and overexpressed and purified the HepA protein. We show in vitro that RNAP preparations have an ATPase activity only in the presence of HepA and that HepA binds core RNAP competitively with the promoter specificity ς70 subunit with a 1:1 stoichiometry and a dissociation constant (K d ) of 75 nm. An E. coli strain with a disruption in the hepA gene shows sensitivity to ultraviolet light.

DNA-dependent RNA polymerase (RNAP) 1 is the central enzyme of transcription and a major target for the regulation of gene expression. The association of a wide array of accessory proteins with RNAP is critical for the regulation of each phase of the transcription cycle: initiation, elongation, and termination. In addition to accessory factors that interact with RNAP to regulate the transcription process, protein-protein interactions couple RNAP to proteins participating in other cellular processes such as DNA repair (1).
During the development of purification procedures for Escherichia coli RNAP, we noticed the consistent co-purification of a 110-kDa polypeptide. The presence of this contaminant through the last step of varying purification procedures suggested that it was a previously unidentified factor specifically associated with the RNAP. Here, we identify the 110-kDa contaminant as the product of the hepA gene, a putative helicase with extensive sequence similarity to the SNF2 family of proteins (2)(3)(4)(5). Some members of this family, which contains proteins from viral, prokaryotic, and eukaryotic species, are DNAdependent ATPases and participate in cellular processes such as transcription regulation or DNA repair. We have cloned the hepA gene and overexpressed and purified the HepA protein.
We show in vitro that RNAP preparations have an ATPase activity only in the presence of HepA and that HepA binds core RNAP competitively with the promoter specificity 70 subunit with a 1:1 stoichiometry and a dissociation constant (K d ) of 75 nM. An E. coli strain with a disruption in the hepA gene shows sensitivity to the DNA damaging agent UV light.

EXPERIMENTAL PROCEDURES
Cloning of the hepA Gene-The hepA gene was PCR amplified from E. coli BL21 (DE3) genomic DNA using the following primers: HEPAleft, 5Ј-GCCGAACACCCATGGCTTTTACACTTGGTC-3Ј; HEPAright, 5Ј-CCATTTTCGGATCCGTTACTGATGCGTTACAACG-3Ј. NcoI and BamHI sites were engineered into HEPAleft and HEPAright, respectively (underlined), to allow digestion of the amplified products and subsequent ligation into the corresponding sites of the T7-based expression vector pET15b (Novagen) to generate pET15b-HepA. The cloning created a mutation in the second amino acid of the protein (from Pro to Ala) as well as a Gln to Arg mutation of the terminal amino acid.
Purification of Overexpressed HepA-E. coli BL21 (DE3) cells were transformed with pET15b-HepA, grown to an A 600 of 0.6, and induced by the addition of IPTG to a final concentration of 1 mM. Induction was allowed to proceed for 4 h. The cells were then harvested by centrifugation and stored at Ϫ80°C. The cells were thawed, resuspended in 40 mM Tris-HCl, pH 7.9, 300 mM KCl, 10 mM EDTA, and lysed by sonication. The lysate was spun to pellet cell debris. The supernatant was collected and poly(ethyleneimine) (PEI) was slowly added to a final concentration of 0.8% (w/v). The PEI pellet was resuspended in TEGD (10 mM Tris-HCl, pH 7.9, 0.1 mM EDTA, 5% glycerol, 1 mM DTT) ϩ 1 M NaCl, eluting HepA. The HepA containing supernatant was then precipitated by slowly adding 35 g of (NH 4 ) 2 SO 4 /100 ml of solution while stirring, incubating for 15 min, and then centrifuging for 45 min at 11,000 ϫ g. The pellet was resuspended in 1 ml of TEGD ϩ 0.5 M NaCl, and the sample was loaded onto a Sephacryl S-300 (Amersham Pharmacia Biotech) gel filtration column. The HepA containing fractions, which were determined by SDS-PAGE and Coomassie staining, were pooled and precipitated again with (NH 4 ) 2 SO 4 as described above. The pellet was resuspended in TEGD and diluted with TEGD until the conductivity was less than TEGD ϩ 0.2 M NaCl. The sample was then loaded onto a Poros HQ (PerSeptive Biosystems) anion exchange column equilibrated in TEGD ϩ 0.2 M NaCl. Proteins were eluted with a linear NaCl gradient from 0.2 to 0.5 M. The peak of HepA eluted at approximately 0.3 M NaCl. At this stage, the protein was Ͼ95% homogeneous as judged by overloaded SDS-PAGE and Coomassie staining. Nevertheless, a contaminating DNase activity was still present, which was effectively removed by gel filtration chromatography on a Superose 6 HR column (Amersham Pharmacia Biotech) equilibrated with 40 mM Tris-HCl, pH 7.9, 100 mM KCl, 10 mM MgCl 2 . Throughout the purification procedure, protein concentrations were determined using Bradford reagent (Bio-Rad) and measuring the absorbance at 595 nm.
Native Gel Shift Binding Assays-50 pmol of HepA protein was added to 25 pmol of core RNAP in 40 mM Tris-HCl, pH 7.9, 100 mM KCl, 10 mM MgCl 2 , 5% glycerol. Samples were left to incubate for 15 min at 37°C. Xylene cyanol was added to 0.25% prior to loading on a 5% polyacrylamide (29.2:0.8 acrylamide:bis) nondenaturing gel (6) in Trisglycine running buffer, pH 8.3. The gels were electrophoresed at 20 -30 mA until the xylene cyanol had migrated about three-fourths the length of the gel. Protein bands were visualized by Coomassie staining. To unambiguously identify the protein contents of the visualized bands, the protein bands of interest were excised from the gel with a scalpel, crushed, and * This work was supported in part by a Burroughs Wellcome Fund Career Award in the Biomedical Sciences (to K. S.) and by grants from the Irma T. Hirschl Trust and a Pew Scholar Award in the Biomedical Sciences (to S. A. D.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ Supported by Rockefeller University Undergraduate Research Fellowships.
For competition binding studies, 50 pmol of HepA protein was added to 25 pmol of core RNAP and incubated as above to form the core RNAP-HepA complex. 50 pmol 70 was then added and incubated for another 15 min at 37°C. Similarly, 50 pmol 70 was incubated with core RNAP to form holoenzyme, which was then incubated with 50 pmol HepA. The samples were then analyzed by the native gel shift assay described above, along with excision of the bands and SDS-PAGE to determine the protein content of the bands.
For the quantitative binding assays to determine the dissociation constant and stoichiometry of the HepA-core RNAP interaction, calf heart protein kinase was used to phosphorylate HepA. HepA (20 g) was incubated with [␥-32 P]ATP and 100 units of calf heart protein kinase (Sigma) in 20 mM Tris-HCl, pH 7.9, 50 mM KCl, 5 mM MgCl 2 , 10 mM DTT. The phosphorylation reaction proceeded for 10 min at 30°C. Unincorporated radioactivity was removed from the reaction mixture by buffer exchange using a Microcon 30 microconcentrator (Amicon) into 40 mM Tris-HCl, pH 7.9, 100 mM KCl, 10 mM MgCl 2 , 5% glycerol. Varying amounts of HepA protein (1-12 pmol) were added to 3 pmol core RNAP and incubated as above. The samples were then analyzed by the native gel shift assay. The bound and free HepA was quantified by a PhosphorImager (Molecular Dynamics Storm).
Construction of hepA Disruption Mutants-Insertion inactivation mutants of the hepA gene were constructed using a modified version of a gene replacement method (7) to create a partial deletion in the hepA locus. The SmaI-digested kanamycin (Km) resistance cassette from pBR⍀km2 (8) was inserted into the end-filled PmlI site of pET15b-HepA using standard molecular biology procedures (9), resulting in a plasmid that carried an interrupted copy of the hepA gene (at codon 304 of 969). The hepA-Km r cassette was then excised from pET15b-HepA with SphI and BspDI and ligated into the SphI and AccI sites of pMAK705 (7) to construct pSD4. E. coli strains NM522 (Invitrogen) and JC7623 (recBC sbcBC; Ref. 10) were then transformed with pSD4 and selected for Km and chloramphenicol (Cm) resistance at 30°C. Single colonies were transferred to Luria Broth (LB) containing both Km and Cm and grown at 44°C until turbid. Dilutions were spread onto prewarmed plates containing both antibiotics and incubated at 44°C to select for cointegrates. The plasmid was then resolved by growing colonies for one cycle at 30°C in the presence of both antibiotics until turbid (overnight) in 100 ml of LB. 100 l of turbid culture was then inoculated into 100 ml of LB containing Km but not Cm, and two cycles of growth were allowed. Single colonies were then selected on media containing Km and grown at 30°C. Cm s -Km r clones were selected by duplicate plating on media containing both antibiotics. Chromosomal insertion into hepA was verified by both PCR analysis as well as Southern analysis.
ATPase Assay-Highly purified fractions of HepA from the Superose 6 column were assayed for their ability to hydrolyze the ␥-phosphate from [␥-32 P]ATP (11). Reactions (10 l) contained 10 M ATP, 3 Ci of [␥-32 P]ATP, 40 mM Tris-HCl, pH 7.9, 100 mM KCl, 10 mM MgCl 2 , and 5 l of each protein fraction. The reactions proceeded at 25°C for 3 h and were terminated by spotting 1 l on a PEI-F cellulose plate (J. T. Baker). The plates were developed in a solution of 0.5 M LiCl, 1 M formic acid and visualized and quantified using a PhosphorImager. ATPase assays were also performed in the presence of 0.5 mg/ml tRNA, doublestranded or single-stranded (boiled) salmon sperm DNA (Life Technologies, Inc.), M13 single-stranded DNA, double-stranded or singlestranded (boiled) pBR322 plasmid DNA, or core RNAP.
For the experiment shown in Fig. 3C, the plasmid pHepA was constructed by cloning the NdeI-HindIII fragment (containing the HepA gene) from pET15b-HepA into the corresponding sites of the polylinker of pTrc99c (22). The resulting plasmid expressed HepA under the control of the IPTG-inducible Trc promoter. RNAP was purified by a standard procedure (12)(13)(14) from the hepA mutant JC7623⌬hepA and also from JC7623⌬hepA/pHepA, and the hepA mutant was transformed with pHepA and induced with 1 mM IPTG. The purified RNAP samples were tested for ATPase activity using the same assay as described above.
UV Sensitivity Assay-The hepA mutants, NM522⌬hepA and JC7623⌬hepA, as well as the parental strains, NM522 and JC7623, were tested for resistance to UV exposure as follows. Cells were grown in 10 ml of LB to an optical density of 0.6 -0.7. In a dark room, cells were exposed to a handheld UV light (254 nM, 4 W, 20 -25 cm distance) in 1-min increments up to 6 min. Dilutions were plated, and the percentage of survival was calculated by comparing colony forming units against control samples not exposed to UV.

RESULTS
A 110-kDa Protein Co-purifies with E. coli RNAP-The initial observation that formed the impetus for this study is illustrated in Fig. 1. In attempting to purify RNAP by a standard procedure (12)(13)(14) from E. coli RL324 (harboring a C-terminal His 6 tag in the chromosomal copy of rpoC, which codes for the RNAP ␤Ј subunit, obtained from R. L. Landick), we noticed nearly stoichiometric amounts of a contaminating protein with a mobility by SDS-PAGE corresponding to about 110 kDa. The contaminating protein was present through the final step of purification in some core RNAP and holoenzyme fractions off an anion exchange column (Fig. 1, fractions 11, 12, 16, and 17). The contaminant was also present through the last step of a different purification procedure utilizing Ni 2ϩ affinity chromatography. The presence of nearly stoichiometric amounts of the 110-kDa contaminant through the final steps of two different purification procedures suggested that this protein interacts with the RNAP itself. The 110-kDa protein was also present after purification of RNAP from E. coli JC7623 (10), the parent strain of RL324 containing the wild-type rpoC gene, eliminating the possibility that the 110-kDa contaminant adventitiously associated with the RNAP from RL324 through the His 6 tag at the C terminus of the ␤Ј subunit.
The 110-kDa Contaminant Is the Product of the HepA Gene-The N-terminal sequence of the 110-kDa contaminant was determined to be XPFTLGQRWISDTESELGL. A data base search revealed a single match to the product of an open reading frame denoted hepA (helicase putative; Refs. [15][16][17]. The inferred amino acid sequence of HepA contained amino acid sequence similarity with motifs I, II, and III of the six "DEAD box" helicase motifs (2). Later, a frameshift in the original DNA sequence was postulated that revealed that the downstream sequences also contained DEAD box helicase motifs V and VI (3). Based on extensive sequence similarity, HepA has already been grouped with the SNF2 family of putative helicases (3,5).
We used PCR methods to clone the hepA gene into a T7based overexpression vector (18). Because of ambiguity in the HepA sequence between helicase motifs III and V (where the frameshift was proposed to occur but its exact location could not be determined; Ref. 3), we sequenced this region of the gene. The sequencing confirmed the presence of a frameshift in the original sequence. The correct sequence gave rise to a peptide of 969 amino acids and a calculated molecular mass of 109,700 Da, consistent with the SDS-PAGE mobility. Between the time our sequencing was completed and this manuscript was written, an updated sequence of this region of the E. coli genome FIG. 1. Highly purified E. coli RNAP contains a 110-kDa contaminant (fractions 16 and 17, labeled with an arrow on the  right). RNAP was prepared using a standard procedure (12)(13)(14). In the final step of purification, RNAP core and holoenzyme were eluted from an anion exchange column by a NaCl gradient. Fractions were analyzed by SDS-PAGE on an 8 -25% Phastgel (Amersham Pharmacia Biotech) and stained with Coomassie Blue. was submitted to GenBank (accession number AE000116 U00096). This sequence exactly matched our sequence.
When induced with IPTG, HepA was expressed to very high levels comprising nearly 50% of total cellular protein. Overexpression of the protein had no obvious toxic effects. Upon cell lysis, the bulk of the overexpressed HepA was found in the soluble fraction. We purified the overexpressed HepA with a procedure similar to that used to purify RNAP (14). In the final step of purification by anion exchange chromatography, the RNAP and HepA bound to the column in buffer containing 0.2 M NaCl. The excess HepA that was not associated with RNAP eluted from the column during a NaCl gradient at about 0.3 M NaCl. The resulting HepA was Ͼ95% homogeneous (based on overloaded, Coomassie-stained SDS gels), but in subsequent experiments it became clear that a substantial DNase activity either was associated with HepA or was contaminating it. We therefore performed an additional step of purification, gel filtration over a Superose 6 (Amersham Pharmacia Biotech) column. This effectively removed the DNase activity from the HepA protein.
HepA Binds Core RNAP but Not Holoenzyme-With the highly purified HepA in hand, we asked whether HepA formed a stable complex with RNAP in vitro, as suggested above. For this purpose, we used a native gel shift assay ( Fig. 2A). When RNAP holoenzyme was mixed with a 2-fold molar excess of purified HepA and analyzed by polyacrylamide gel electrophoresis under nondenaturing conditions, bands corresponding to free HepA and free RNAP holoenzyme were observed ( Fig.  2A, lane 1). Because mobility on the native gel is determined by both molecular weight and charge, the band in Fig. 2A corresponding to RNAP holoenzyme could conceivably contain a holoenzyme-HepA complex with the same mobility as holoenzyme. Therefore, we confirmed the protein components of the bands labeled in Fig. 2A by excising them and analyzing their contents by SDS-PAGE (Fig. 2B). By this method, the band labeled A in Fig. 2A contains only ␤Ј, ␤, 70 , ␣, and , the components of holoenzyme (Fig. 2B, lane A). Core RNAP alone gave rise to two bands on the native gel ( Fig. 2A, lane 5), likely because of the presence of core RNAP monomers and dimers in equilibrium (14). A mixture of HepA and core RNAP (2:1 molar ratio) yielded a band corresponding to free HepA and a band distinct from the bands observed for core RNAP alone ( Fig. 2A,  lane 4, band C). This distinct band contained ␤Ј, ␤, ␣, and (the components of core RNAP) and an apparently stoichiometric amount of HepA, based on the intensity of the Coomassie stain (Fig. 2B, lane C). Thus, we conclude that HepA forms a stable complex with core RNAP but not holoenzyme.
Additional experiments were conducted in which preformed complexes of core RNAP/HepA were challenged with an equimolar amount of 70 or RNAP holoenzyme (core RNAP/ 70 ) was challenged with an equimolar amount of HepA. 70 effectively displaced HepA from core RNAP, whereas HepA was unable to displace 70 (data not shown).
We adventitiously found that treatment of purified HepA with [␥-32 P]ATP and calf heart protein kinase resulted in covalent labeling of HepA with [ 32 P]. We took advantage of this ability to radioactively label HepA to quantitate its interaction with core RNAP. Increasing amounts of labeled HepA were mixed with a constant amount of core RNAP, incubated for 15 min at 37°C to form a complex, and then analyzed by the native gel shift assay (Fig. 2C). The amounts of free HepA and HepA associated with RNAP (and therefore shifted to the lower mobility band) were quantitated by PhosphorImager analysis of the gels. A Scatchard analysis of the results (Fig. 2D) revealed that HepA interacts with core RNAP in a 1:1 complex with a dissociation constant (K d ) of 75 nM. The finding that HepA binds core RNAP and not holoenzyme is not in contradiction to our earlier finding that HepA contaminated purified fractions of both core RNAP and holoenzyme off the final MonoQ column fractionation (Fig. 1). During the purification of overexpressed, free HepA, we found that it eluted from the MonoQ column at about 0.3 M NaCl, about the same as core RNAP (13). Therefore, we believe the HepA found in core RNAP fractions (Fig. 1, lanes 11 and 12) corresponds to the unbound fraction of HepA, whereas the HepA found in holoenzyme fractions (Fig. 1, lanes 16 and 17) corresponds to HepA in the core RNAP-HepA complex, which co-elutes roughly with holoenzyme. This is possible because the molar amount of 70 is roughly half that of the RNAP. Based on the Coomassie-stained bands, it appears that roughly 60% of the HepA is bound to RNAP, whereas about 40% is found in the unbound fraction.
Insertion Inactivation of HepA Results in Sensitivity to DNA Damage-To assess the role of HepA in cellular processes, we constructed an insertion of a Km r cassette at codon 304 (between helicase motifs II and III) in the genomic hepA gene of two separate strains of E. coli, NM522, and JC7623 (recBC sbcBC; Ref. 10). Insertion of the Km r cassette at this position of the hepA gene results in a predicted protein product less than one-third the length of full-length HepA and containing only two of the six helicase motifs (I and II). Thus, the normal function of HepA was undoubtedly disrupted.
The two E. coli strains with the disrupted hepA gene (JC7623⌬hepA and NM522⌬hepA) did not exhibit any obvious growth phenotypes over a temperature range of 25-45°C. Because of the role, or suspected role, of many SNF2 family members in various DNA repair processes, we tested JC7623⌬hepA and NM522⌬hepA for their ability to survive exposure to UV light, which is known to cause DNA damage. An example of the results is shown in Fig. 3, where survival of JC7623 and JC7623⌬hepA to increasing times of UV exposure is compared. The results, tabulated in Table I, show clearly that the hepA gene disruption in JC7623⌬hepA and NM522-⌬hepA results in significantly reduced survival to UV exposure compared with the parent strains JC7623 and NM522.
ATPase Activity Associated with HepA-Because several members of the SNF2 family have been shown to be DNA-dependent ATPases (5), we tested fractions of the highly purified HepA from the Superose 6 gel filtration column for ATPase activity. Ability to hydrolyze the ␥-phosphate from [␥-32 P]ATP was monitored by thin layer chromatography (11), revealing an ATPase activity above background (Fig. 4A). The ATPase activity in each fraction was nearly exactly proportional to the protein concentration in the fraction (Fig. 4B).
To rule out the possibility that this weak ATPase activity arose from a contaminant of the purified HepA preparation, we compared the ATPase activity of purified RNAP from JC7623⌬hepA and JC7623⌬hepA/pHepA (JC7623⌬hepA transformed with a plasmid expressing HepA under the control of the Trc promoter). As expected, the purified RNAP from JC7623⌬hepA/pHepA induced with IPTG contained HepA (Fig.   FIG. 3. Disruption of the hepA gene decreased survival to UV exposure. The graph shows a log plot of (% survival) versus time of UV exposure for two E. coli strains, JC7623 (f) and the same strain with a disruption in the hepA gene, JC7623⌬hepA (OE). Each data point represents the mean of five separate experiments. The error bars show the standard deviation of each data point. The best-fit lines are also shown.

NM522
Ϫ0.14 NM522⌬hepA Ϫ0 4C, lane 2), whereas RNAP from JC7623⌬hepA did not (Fig.  4C, lane 3). An ATPase assay of these same samples indicated that RNAP from JC7623⌬hepA had ATPase activity similar to a background sample with no added protein (Fig. 4C, lanes 6  and 7), whereas RNAP from JC7623⌬hepA/pHepA had ATPase activity well above background (Fig. 4C, lane 5). Thus, we conclude that RNAP preparations have an ATPase activity only in the presence of HepA. This ATPase activity was not DNA-or RNA-dependent and was not stimulated by the addition of tRNA, double-stranded DNA, single-stranded DNA, or core RNAP (data not shown). DISCUSSION We have identified HepA, an E. coli protein that shares extensive sequence homology with the SNF2 family of putative helicases (3,5,17), as an RNAP-associated factor. We cloned and purified HepA and showed that RNAP preparations have an ATPase activity only in the presence of HepA and that HepA associates with core RNAP in vitro (but not holoenzyme). We disrupted the hepA gene in E. coli, resulting in a phenotype displaying sensitivity to UV exposure The SNF2 family includes proteins from viral, prokaryotic, and eukaryotic species with roles in cellular processes such as cell cycle control (STH1), transcriptional regulation and chromatin remodeling (ATR-X, BRM, hBRM, MOT1, ISW1, and SNF2), nucleotide excision repair (RAD16 and ERCC6), mitotic recombination (RAD54), and other types of DNA repair (RAD5). The family also includes many proteins with no known function.
HepA is predicted to be an ATPase based on its extensive sequence similarity with other ATPases, and thus the ATPase activity associated with RNAP preparations only in the presence of HepA is likely to belong to HepA itself. The results of our experiments, however, do not rule out the possibility that the ATPase activity is associated with another protein (perhaps the RNAP itself) and that this activity is greatly stimulated by HepA. In contrast to SNF2 family members that have been shown to have ATPase activity (Saccharomyces cerevisiae SNF2 and MOT1, human HIP116A); however, this ATPase activity does not appear to be DNA-dependent.
The close relationship between the SNF2 family of proteins and known helicases (2) led us to test the purified HepA protein for helicase activity on various DNA and RNA substrates. Helicase activity has not been demonstrated for any SNF2 family member, and we were unable to detect any activity for HepA (data not shown).
Because of the observed association between HepA and core RNAP, we also tested the effect of purified HepA protein on various in vitro transcription assays. We tested the effect of HepA on abortive initiation (19) by 70 holoenzyme at the T7 A1 promoter. We also formed ternary elongation complexes containing a 20-mer transcript on the T7 A1 tr2 transcription unit (20), added HepA, and then added nucleotides to initiate transcription elongation. We then examined the effect of HepA on transcription pausing, on the overall transcription elongation rate, and on termination at tr2. Finally, we formed the 20-mer ternary complexes on the T7 A1 tr2 transcription unit and then added ATP and HepA and tested for displacement of the ternary complexes. In all of these investigations, we could not observe any effect of the purified HepA protein on abortive initiation, transcription elongation, or termination or on the stability of the stalled ternary complexes (data not shown).
We constructed insertion inactivation mutants of hepA to obtain clues to the role HepA plays in cellular processes. The hepA disruption mutants were sensitive to UV exposure, suggesting that they were defective in some DNA repair process. The finding that HepA associates with RNAP links HepA with transcription, whereas the data from the hepA gene disruptions link HepA with DNA repair.
It is interesting to note that the original observation of HepA co-purification with RNAP came from E. coli RL324 (harboring a C-terminal His 6 tag in the chromosomal copy of rpoC, which codes for the RNAP ␤Ј subunit, obtained from R. L. Landick), and subsequently from E. coli JC7623 (10), the parent strain of RL324 containing the wild-type rpoC gene, eliminating the possibility that HepA adventitiously associated with the RNAP from RL324 through the His 6 tag at the C terminus of the ␤Ј subunit. Although we have found HepA co-purification with RNAP from other E. coli strains, the amounts of HepA associated with RNAP in these two strains is always substantially greater, suggesting that expression of HepA is up-regulated in these strains. Both RL324 and its parent JC7623 are recBC sbcBC mutants. These mutations in the RecBCD enzyme complex are necessary for the efficient transformation of linear DNA into E. coli and were thus used for the construction of RL324. The RecBCD enzyme complex plays important roles in both recombination and DNA repair pathways. Thus, it is interesting to speculate that the possible disruption of RecBCD function may be compensated by increased expression of HepA, which we have linked to DNA repair.