A Domain of RecC Required for Assembly of the Regulatory RecD Subunit Into the Escherichia coli RecBCD Holoenzyme

Corresponding author: Gerald R. Smith, 1100 Fairview Ave. North, P.O. Box 19024, Seattle, WA 98109., gsmith{at}fhcrc.org (E-mail)

Communicating editor: L. S. SYMINGTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The heterotrimeric RecBCD enzyme of Escherichia coli is required for the major pathway of double-strand DNA break repair and genetic exchange. Assembled as a heterotrimer, the enzyme has potent nuclease and helicase activity. Analysis of recC nonsense and deletion mutations revealed that the C terminus of RecC is required for assembly of the RecD subunit into RecBCD holoenzyme but not for recombination proficiency; the phenotype of these mutations mimics that of recD deletion mutations. Partial proteolysis of purified RecC polypeptide yielded a C-terminal fragment that corresponds to the RecD-interaction domain. RecD is essential for nuclease activity, regulation by the recombination hotspot Chi, and high affinity for DNA ends. The RecC-RecD interface thus appears critical for the regulation of RecBCD enzyme via the assembly and, we propose, disassembly or conformational change of the RecD subunit.

DNA repair and recombination are highly regulated processes that require the modulation of degradative and recombinogenic activities of multiple enzymes that act on DNA. Such regulation can alter the distribution of recombination around special sites on the chromosome. For example, the RecBCD enzyme of Escherichia coli is regulated by complex interactions between its three subunits and a Chi site (GCTGGTGG), a DNA sequence near which recombination occurs at high frequency (TAYLOR 1988 ; KOWALCZYKOWSKI et al. 1994 ; SMITH 2001 ). The enzyme is a heterotrimer with a mass of 330 kD, composed of one copy of the products of the recB, recC, and recD genes (TAYLOR and SMITH 1995 ), and is required for the major pathway of homologous recombination and DNA repair involving linear double-stranded (ds) DNA. RecBCD enzyme is an ATP-dependent ds and single-stranded (ss) exonuclease, an ss endonuclease, a DNA helicase, and a DNA-dependent ATPase. Full enzymatic activity requires all three enzyme subunits; in some cases one or two subunits containing identified domains demonstrate weaker enzymatic activity (BOEHMER and EMMERSON 1992 ; KORANGY and JULIN 1992 , KORANGY and JULIN 1993 ; YU et al. 1998B ). ATPase domains have been identified in RecB and RecD (HICKSON et al. 1985 ; FINCH et al. 1986A ; JULIN and LEHMAN 1987 ), and at least part of the nuclease domain is in the C-terminal portion of RecB (YU et al. 1998A , YU et al. 1998B ; WANG et al. 2000 ).

The structure and function of RecBCD enzyme is regulated by Chi sites, hotspots of homologous recombination at which a 3' ssDNA end is made (PONTICELLI et al. 1985 ; DIXON and KOWALCZYKOWSKI 1993 ) and loaded with RecA protein (ANDERSON and KOWALCZYKOWSKI 1997 ). The RecA·ssDNA filament then undergoes pairing and strand exchange with a homologous chromosome (KOWALCZYKOWSKI 2000 ). The RecD subunit plays at least two roles in the RecBCD enzyme-Chi interaction. First, the RecD subunit is required, along with a nuclease domain in RecB, for nuclease activity (CHAUDHURY and SMITH 1984A ; YU et al. 1998A , YU et al. 1998B ); this activity is required to form the recombinogenic 3' ssDNA end at Chi. Second, RecD interferes with a RecA-loading domain in RecBCD enzyme (AMUNDSEN et al. 2000 ). RecBC enzyme (without the RecD subunit) loads RecA on the DNA constitutively during DNA unwinding (CHURCHILL et al. 1999 ), whereas RecBCD enzyme does not load RecA until the enzyme acts at Chi (ANDERSON and KOWALCZYKOWSKI 1997 ). RecD and a nuclease-dependent signal thus play a role in the regulation of RecA loading (ANDERSON et al. 1999 ; AMUNDSEN et al. 2000 ). As a consequence of RecBCD enzyme regulation, both the DNA substrate and RecBCD enzyme are changed by an interaction at Chi (TAYLOR and SMITH 1992 , TAYLOR and SMITH 1999 ). A RecBCD enzyme molecule that acts at one Chi site cannot act at a second Chi site on the same DNA molecule, although DNA unwinding continues to the end of the molecule (TAYLOR and SMITH 1992 ). After acting at Chi, RecBCD enzyme disassembles into its three subunits, rendering the enzyme inactive on another DNA molecule (TAYLOR and SMITH 1999 ). Understanding the assembly and disassembly of RecBCD enzyme is thus important in understanding regulation of the enzyme.

Analysis of recBCD mutants has established the role of certain RecBCD enzyme subunits and activities in recombination proficiency. Null mutations in recB or recC eliminate recombination proficiency and all enzymatic functions (Rec^- Nuc^-; HOWARD-FLANDERS and THERIOT 1966 ; WILLETTS et al. 1969 ; WILLETTS and MOUNT 1969 ), whereas null mutations in recD do not affect recombination proficiency but do eliminate nuclease activity and Chi activity (Rec⁺ Nuc^-; i.e., the phenotype; CHAUDHURY and SMITH 1984A ; AMUNDSEN et al. 1986 ). The phenotype is consistent with the DNA-unwinding and RecA-loading activity of RecBC enzyme as described above. Two mutations in recC also have the phenotype (CHAUDHURY and SMITH 1984A ; this work). These special recC mutants could be explained by a failure to assemble RecD into RecBCD holoenzyme. We show here that the C-terminal portion of RecC is altered in these mutants and is indeed required for RecD assembly into holoenzyme. Collectively, our data identify a 35-kD domain of RecC that is required for RecD assembly and consequently is essential for the regulation of RecBCD enzyme by Chi.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Bacterial strains, plasmids, and genotype designations:
All strains and plasmids are listed in Table 1 with their genotypes and sources. For visual clarity, allele numbers are expressed as superscripts when more than one recBCD gene is designated. The corresponding polypeptide designation is also expressed as a superscript.

Mutant isolation:
The isolation of the recC1010 mutant has been described (CHAUDHURY and SMITH 1984A ). The recC1041 mutant was produced by treating strain S927 (F'15 thyA⁺ recBCD⁺ argA⁺; SCHULTZ et al. 1983 ) with nitrosoguanidine (Sigma, St. Louis) as described by AMUNDSEN et al. 1990 . The mutagenized F' factors were transferred to strain V186, and rec1041 was identified by altered phage sensitivity (SCHULTZ et al. 1983 ), indicating an alteration in RecBCD enzyme activity. Complementation analysis, performed as described by AMUNDSEN et al. 1990 , showed that the rec1041 mutation was in recC (data not shown). An 18.5-kb BamHI fragment containing the recBC¹⁰⁴¹D genes was cloned into the BamHI site of pBR322 as described by PONTICELLI et al. 1985 , to create pSA160.

RecBCD enzyme purification and detection:
RecBC¹⁰⁴¹ enzyme was purified from 3 liters of V68 (recC73) carrying pSA160 (recBC¹⁰⁴¹D) as described by AMUNDSEN et al. 2000 . Purification was through HiTrapQ, Sephacryl S-300, and HiTrap heparin (all from Amersham Pharmacia). Enzymes were analyzed by native (5% polyacrylamide in 50 mM MOPS-KOH, pH 7.0, and 1 mM EDTA) or SDS polyacrylamide (3–8% polyacrylamide in 50 mM Tris-acetate buffer, pH 8.25, with 0.1% SDS) gel electrophoresis as described by TAYLOR and SMITH 1999 . The concentration of RecBC¹⁰⁴¹ enzyme was determined by comparing the amount of Coomassie-stained protein material on native and SDS polyacrylamide gels to known amounts of RecBC and RecBCD enzyme. Western blots were probed with mouse monoclonal antibodies specific for RecB, RecC, or RecD (TAYLOR and SMITH 1999 ). RecBC enzyme was prepared by mixing purified RecB and RecC polypeptides (BOEHMER and EMMERSON 1991 ) and further purified on a HiPrep Sephacryl S-300HR column (Amersham Pharmacia). RecBCD mutant enzymes were also analyzed in cell extracts prepared as described by TOMIZAWA and OGAWA 1972 except that cultures were grown in Terrific broth (Sigma).

RecBCD enzyme reaction conditions and DNA substrates:
RecBCD enzyme was assayed for ATP-dependent solubilization of ³H-labeled dsDNA (EICHLER and LEHMAN 1977 ; AMUNDSEN et al. 1990 ). Assays for DNA unwinding and RecA loading (CHURCHILL et al. 1999 ) used plasmid pBR322 DNA digested with HindIII. Substrate preparation and reaction conditions were as reported by AMUNDSEN et al. 2000 .

Protease cleavage of RecC polypeptide:
Purified RecC polypeptide (20 µg) in 80 µl of protease digestion buffer [20 mM KPO₄, pH 6.8, 5 mM dithiothreitol (DTT), 10% glycerol] was digested at 25° with 0.9 ng of subtilisin (Sigma) or 1.9 ng of trypsin (Sigma) for the times indicated. At each time point an 8-µl sample was removed and phenylmethylsulfonyl fluoride was added to 10 mM. Samples were analyzed by electrophoresis through a 4–12% Bis-Tris SDS gel (Invitrogen, San Diego) in 50 mM MOPS-SDS buffer (Invitrogen). The digestion products were visualized by staining (Fast stain, Zoion Research, Allston, MA). The 35-kD product of trypsin cleavage was transferred to a PVDF membrane (Millipore, Bedford, MA) by electroblotting in CAPS buffer (MATSUDAIRA 1987 ) for N-terminal sequence analysis. Proteins were sequenced on an Applied Biosystems (Foster City, CA) model 492 N-terminal protein sequencer using Edman degradation reactions.

Subcloning and exonuclease III deletion analysis of recC:
The recC gene was cloned by DNA amplification with Expand DNA polymerase (Roche, Indianapolis) from pDWS2 (recBCD⁺; PONTICELLI et al. 1985 ) template DNA. The upstream primer (CGCGGATCCGCGGGGTACCAACAGCTCTGGCGGCATGGCTGG), beginning 195 nucleotides upstream of the recC start codon, and the downstream primer (CGCGGATCCCCGGAATTCGCGGATAGATTGCGCAATTTTTATACAG), with 14 nucleotides following the recC stop codon, both contained a BamHI recognition sequence. The product of DNA amplification was digested with BamHI and ligated into the BamHI site of Bluescript II KS (Stratagene, La Jolla, CA), yielding pSA125. Exonuclease III deletions were made with the Erase-a-Base system (Promega, Madison, WI) after digesting pSA125 with ApaI and XhoI. The extent of each deletion was determined by sequencing plasmid DNA. The partially deleted plasmids were introduced into strain V68 (recC73 null mutant) and assayed for recombination proficiency, nuclease activity, and enzyme subunit assembly (Fig 4 and Fig 5).

View larger version (19K): In this window In a new window Download PPT slide   Figure 1. The recC mutations map near the C terminus of RecC. The genetic map of the thyA-argA region of the E. coli chromosome shows the position and type of certain recC mutations and the accompanying predicted codon changes (data not shown; recC73 was reported by ARNOLD et al. 2000 ). The region containing a protease-hypersensitive cleavage site (codons 805–819) is indicated (*). Nucleotide numbering begins with the A of the ATG start codon (FINCH et al. 1986B ).

View larger version (36K): In this window In a new window Download PPT slide   Figure 2. RecC mutants do not assemble RecD into RecBCD enzyme heterotrimer. Extracts were from late log phase cultures of strain V330 [(recC-argA)234] containing derivatives of plasmid pDWS2 (recBCD), pAC2 (recBC1010D; AMUNDSEN et al. 1986 ), or pSA160 (recBC1041D). Purified RecC, RecBC, and RecBCD enzyme (40 ng each) were run as markers. Forms of RecBCD enzyme contained in 7.5 or 15 µg of extract protein were separated on native or SDS polyacrylamide gels and detected by Western analysis using the indicated monoclonal antibodies. The RecC polypeptide shown in lane 3 of part B had been stored at 4° for several months and contained an apparent protease cleavage product (see RESULTS).

View larger version (37K): In this window In a new window Download PPT slide   Figure 3. RecBC1041 enzyme unwinds DNA and loads RecA protein. RecBC and RecBC1041 enzyme were assayed using 5'-32P-labeled pBR322 DNA as described (CHURCHILL et al. 1999 ; AMUNDSEN et al. 2000 ). The DNA substrate (4.7 nM) and the indicated amount of RecBC or mutant enzyme were incubated at 37° for 5 min with or without RecA protein as indicated. An aliquot was removed for analy-sis (0 min Exo I). Exo-nuclease I was added to the remaining sample and incubation continued for 7 and 12 min. After addition of SDS to dislodge RecA protein, the products of the reaction were analyzed by electrophoresis in a 1% agarose gel. The positions of dsDNA substrate (DS, lane 1) and unwound ssDNA (SS, boiled, lane 2) are shown.

View larger version (28K): In this window In a new window Download PPT slide   Figure 4. Deletions of the C terminus of recC produce the (RecD-) phenotype. Open bars indicate the extent of the RecC polypeptide remaining in the exonuclease III-generated deletions. The extents of RecC polypeptide expressed by recC1041 (nonsense) and recC2723 (N-terminal expression subclone) are shown for comparison. The phenotype of each derivative was determined in genetic assays as described in Table 2 for recombination proficiency (Hfr recombination) and intracellular nuclease activity (efficiency of plating of phage T4 gene 2- mutants). (+) The presence of nuclease activity, which resulted in an efficiency of plating of phage T4 2- of <1 x 10-4. (-) The lack of nuclease activity resulting in an efficiency of plating of >0.8.

View larger version (38K): In this window In a new window Download PPT slide   Figure 5. C-terminal deletions of recC fail to assemble RecD into holoenzyme. Extracts were prepared as described in Fig 2 from V330 [(recC-argA)234] containing plasmids expressing full-length RecC (A, pSA125; B, pSA161) or the indicated C-terminal deletions. Purified RecB polypeptide or extract containing RecD polypeptide, or both, were added to the RecC extract as indicated, and assembly of heterodimer and heterotrimer forms of RecBCD enzyme were monitored on native polyacrylamide gels as in Fig 2A.

Fragments of RecC flanking the trypsin-hypersensitive site (see RESULTS) were expressed from the trc promoter of pSE380 (Invitrogen). recC⁺ (codons 1–1122, pSA161), recC2723 (codons 1–804, pSA162), and recC2724 (codons 805–1122, pSA163) were cloned by DNA amplification with Expand DNA polymerase (Roche) from pDWS2 template DNA (details available upon request).

RecBCD enzyme assembly following subunit mixing:
Purified RecB and RecC subunits were prepared as described by BOEHMER and EMMERSON 1991 . Other subunits [RecD from pB100 and pNH52 (BOEHMER and EMMERSON 1991 ) and RecC from pSA125 and deletion derivatives] were provided in extracts of V330 [(recC-argA)234] carrying appropriate plasmids. The approximate concentration of enzyme subunits in cell extracts was determined by fractionating extracts and known amounts of RecBCD enzyme on denaturing polyacrylamide gels and comparing the signal intensities following Western blot analysis using RecD or RecC monoclonal antibodies. For the enzyme assembly assay, RecBCD enzyme subunits (8 nM RecD, 32 nM RecC, 32 nM RecB) were combined in R buffer (20 mM KPO₄, pH 7.6, 0.1 mM EDTA, 50 mM DTT, 100 mM NaCl, 10% glycerol, 1 mg/ml BSA) and incubated overnight at room temperature. Samples were analyzed by electrophoresis on native polyacrylamide gels (see above) to assess the state of enzyme assembly.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

recC mutants, like recD nonsense mutants, are recombination proficient but lack nuclease activity:
During a search for novel recBCD mutants we found a candidate, recC1041, that appeared to have the phenotype (Rec⁺ Nuc^-; CHAUDHURY and SMITH 1984A ; AMUNDSEN et al. 1986 ; MATERIALS AND METHODS). To determine if the candidate was indeed a mutant, we first measured recombination following Hfr conjugation (Table 2). We compared recombinant frequencies of recC1041 with those of two known mutants, recC1010 and recD1013, a nonsense mutation in the fourth codon of recD (data not shown; CHAUDHURY and SMITH 1984A ). The two special recC mutants and the recD1013 nonsense mutant were almost as recombination proficient as the wild-type strain (Table 2) and produced 100 times more recombinants than did the null (recC73) mutant control. Recombination proficiency was also reflected by the resistance to the DNA-damaging agent mitomycin C manifested by the recC, recD, and wild-type strains relative to the recB21 null mutant (Table 2; as noted above recB and recC null mutants have identical phenotypes).

View this table: In this window In a new window   Table 2. recC1010 and recC1041 mutants have the same phenotype (Rec+ Nuc-, ) as recD nonsense mutants

To determine if the recC1041 mutant lacked nuclease activity, we measured this activity in mutant and wild-type strains directly in extracts and indirectly in cells. Extracts of the recC1010 or recC1041 mutants, like the recD1013 mutant, had <2% of the ATP-dependent exonuclease activity of the wild-type (recBCD⁺) strain (Table 2). An assay that detects RecBCD enzyme nuclease activity inside the cell confirmed these observations. The phage T4 gene 2 product binds to the ends of T4 DNA, thereby protecting the injected DNA from degradation by RecBCD enzyme (OLIVER and GOLDBERG 1977 ). Consequently, T4 gene 2^- mutants make plaques only on strains lacking RecBCD enzyme nuclease activity. The titer of T4 2^- phage on the recC1010 and recC1041 strains, like that on the recD nonsense mutant, was 10⁶ times higher than that on the wild-type strain, reflecting the absence of dsDNA exonuclease activity in both recC mutants and the recD mutant. These data show that recC1010 and recC1041 are mutants: they are recombination proficient but lack nuclease activity (Rec⁺ Nuc^-).

The recC mutants have the same phenotype as a recD nonsense mutant (Table 2), suggesting that RecD is not active in the recC mutants. If a region of RecC interacts with RecD, the recC mutations might cluster close together and eliminate RecD association with RecBCD enzyme. We sequenced the recC genes from these mutants and found that the two mutations were 191 bp apart near the C terminus of recC (Fig 1; data not shown). recC1010 is a missense mutation at codon 905 of 1122 recC codons, and recC1041 is a nonsense mutation at codon 841. These results indicate that the region of RecC missing in RecC1041 (amino acids 841–1122) is not required for recombination proficiency but, as shown next, is needed for assembly of RecD into holoenzyme.

recC mutants fail to assemble RecBCD holoenzyme:
We hypothesized that the recC phenotype could be explained by the failure to assemble RecD into holoenzyme, functionally yielding RecBC enzyme inside the cell. We therefore examined the state of RecBCD enzyme in extracts of wild-type and recC mutant strains. Native polyacrylamide gel electrophoresis resolved purified RecC, RecBC, and RecBCD holoenzyme, as revealed by Western blots probed with a monoclonal antibody specific for RecC (Fig 2A, lanes 1–3). The wild-type recBCD⁺ extract contained RecBCD holoenzyme, free RecC, and a small amount of RecBC (lanes 4–5). In contrast, extracts of recC1010 and recC1041 contained RecBC and RecC but no detectable RecBCD holoenzyme (lanes 8–11). We show below that these extracts contained all three polypeptides. These results show that the recC mutants fail to assemble RecD into the enzyme complex.

Analysis of RecC polypeptides on denaturing (SDS) polyacrylamide gels confirmed that the recC1041 nonsense mutation resulted in the production of a truncated form of RecC (Fig 2B, lanes 11–12; data not shown). The RecC1041 polypeptide had a molecular mass of 95 kD, as predicted by the position of the nonsense mutation in the gene (at codon 841; see above). Wild-type (recBCD⁺) and recC1010 extracts (lanes 7–10) contained full-length RecC polypeptide that comigrated with purified RecC polypeptide (lane 4), or RecC in RecBC enzyme and RecBCD holoenzyme (lanes 1 and 2). The purified RecC polypeptide in lane 3 of Fig 2B had been stored at 4° for several months and contained an apparent protease cleavage product whose size was also 95 kD (see also Fig 6; this cleavage product is discussed further below).

View larger version (60K): In this window In a new window Download PPT slide   Figure 6. Physical evidence for two RecC domains. Purified RecC polypeptide (2 µg per lane) was digested with subtilisin (lanes 2–7) or trypsin (lanes 9–14) for the times indicated. Lanes 1 and 15 contained size markers (Bio-Rad) and lane 8 contained RecC polypeptide that was stored for several months at 4°. The digestion products were analyzed by gel electrophoresis and visualized by staining (Fast stain, Zoion Research).

We determined that extracts of the recC mutants contained RecD polypeptide even though it was not assembled into RecBCD holoenzyme. Western blots of SDS polyacrylamide gels were probed with a monoclonal antibody specific for RecD (Fig 2C). RecD polypeptide was detected in purified RecBCD enzyme (lane 2) and extracts of recBCD⁺, recC1010, and recC1041 strains (lanes 7–12). A similar analysis performed with a RecB monoclonal antibody demonstrated that all extracts contained RecB polypeptide as well (data not shown). These data show that the recC mutants produced RecD but failed to assemble it into holoenzyme and suggest that a C-terminal domain of RecC is essential for this process.

Our view of RecC polypeptide now includes two general regions (Fig 1). The first, the N-terminal region, is apparently sufficient, with RecB polypeptide, to mediate recombination proficiency (Table 2). The second, the C-terminal region, missing in RecC1041 and altered by a single amino acid change in RecC1010, is essential for RecBCD holoenzyme assembly (Fig 2).

Purified RecBC¹⁰⁴¹ enzyme is a heterodimer, unwinds DNA, and loads RecA protein:
RecBCD enzyme must facilitate the loading of RecA protein to support recombination (AMUNDSEN et al. 2000 ; ARNOLD and KOWALCZYKOWSKI 2000 ). The recombination proficiency of recD mutants has been attributed to the ability of the RecBC enzyme (missing the RecD subunit) to unwind DNA and facilitate the loading of RecA protein onto ssDNA that is produced during DNA unwinding (CHURCHILL et al. 1999 ). Because recC1041 is recombination-proficient (Table 2), we expected purified RecBC¹⁰⁴¹ enzyme to have the enzymatic activities associated with RecBC enzyme.

We first examined the state of assembly of purified enzyme from the recC1041 strain. After electrophoresis on a native polyacrylamide gel, a Western blot was probed with a RecC monoclonal antibody and revealed that the preparation contained RecBC¹⁰⁴¹ heterodimers but not RecBC¹⁰⁴¹D heterotrimers (data not shown). Thus, RecBC¹⁰⁴¹ enzyme was a heterodimer in crude extracts (Fig 2A) and after purification, indicating a defect in enzyme assembly in both situations.

We next compared the DNA-unwinding and RecA-loading activities of purified RecBC enzyme and RecBC¹⁰⁴¹ enzyme (Fig 3). As shown previously (CHURCHILL et al. 1999 ), RecBC enzyme unwound the dsDNA substrate, producing ssDNA (lanes 3, 6, 9, and 12). Loading of RecA protein on the 3' end of ssDNA protected the DNA from exonuclease I digestion (lanes 4, 5, 10, and 11); this protection required RecA to be present in the reaction (lanes 7, 8, 13, and 14). DNA produced by boiling was not protected in the presence of subsequently added RecA protein and RecBC (data not shown), as expected from the requirement that RecBC enzyme actively unwind the DNA to load RecA (CHURCHILL et al. 1999 ). As expected from the recombination proficiency of recC1041, we detected DNA-unwinding and RecA-loading activity in purified RecBC¹⁰⁴¹ enzyme (lanes 15, 18, 21, and 24). Unwound ssDNA was protected from exonuclease I only in the presence of RecA (compare lanes 16 and 17 and 22 and 23 with lanes 19 and 20 and 25 and 26). The activity of purified RecBC¹⁰⁴¹ enzyme in this assay, although weaker than that of RecBC enzyme, is qualitatively consistent with the recombination proficiency observed in genetic assays of recC mutants. There is a similar quantitative difference in activity between RecBCD enzyme and RecBC enzyme (CHURCHILL et al. 1999 ; AMUNDSEN et al. 2000 ). RecBC enzyme has much less affinity for DNA than does RecBCD enzyme (data not shown) yet is sufficiently active to support recombination proficiency in the cell (CHAUDHURY and SMITH 1984A ).

We next tested whether purified RecBC¹⁰⁴¹ enzyme had nuclease activity. In accord with the assays of unfractionated extracts (Table 2), purified RecBC¹⁰⁴¹ enzyme had <0.02% of the nuclease activity of RecBCD enzyme (1.9 x 10⁵ ds exonuclease units per milligram of RecBCD polypeptides). Thus, the truncated RecC1041 polypeptide, combined with RecB, had DNA-unwinding and RecA-loading activities but not dsDNA exonuclease activity. These activities are qualitatively the same as those of RecBC enzyme with full-length RecC polypeptide.

C-terminal deletions of recC also have the phenotype and fail to assemble RecD:
Since recC1041 () is a nonsense mutation, we reasoned that C-terminal recC deletion mutations would also have the phenotype (Rec⁺ Nuc^-) and would further define regions of RecC required for recombination proficiency and RecBCD enzyme assembly. As predicted, exonuclease III-generated deletions of the C terminus of recC, ranging from 38 to 332 codons, produced the phenotype (Fig 4). Thus, 50 codons toward the N terminus of RecC beyond recC1041 could be deleted without affecting recombination proficiency. In contrast, deletion of >444 recC codons produced a null phenotype (Rec^- Nuc^-), indicating the loss of an essential function such as DNA binding, DNA unwinding, or RecA loading when 678 codons of recC remain.

In summary, the analysis of recC C-terminal deletions identified mutations with three distinct phenotypes: wild-type (expressing codons from 1 to 1118), (expressing codons from 1 to 790–1084), and null (expressing codons from 1 to 456–678). Taken together, these results show that codons 1–790 of RecC are sufficient to confer a recombination-proficient phenotype: the C terminus of RecC is not required for recombination proficiency.

We next determined whether the RecC deletion polypeptides, like the RecC1041 nonsense fragment, failed to assemble RecD into holoenzyme. We monitored RecD assembly into holoenzyme after mixing purified RecB polypeptide with extracts containing RecD and selected RecC deletion polypeptides (Fig 5). Subunit monomers, heterodimers, and heterotrimers were detected on Western blots of native polyacrylamide gels probed with a RecC monoclonal antibody. RecBCD enzyme heterotrimers were detected only in the presence of RecB, RecD, and full-length RecC polypeptide. Expression of 1084, 981, or 889 codons of recC allowed RecB and RecC association but not RecD assembly into holoenzyme (Fig 5A). The absence of RecD in the enzymes indicates that the missing C termini of RecC were required for enzyme assembly. The extent of RecBC heterodimer formation varied, being greater for more nearly full-length polypeptides than for the shorter versions. Nevertheless, the corresponding recC deletion strains were recombination proficient (Fig 4), indicating that these RecBC enzyme derivatives were able to unwind DNA and load RecA protein.

RecC polypeptides with amino acids 1–790 or 1–804 did not form detectable amounts of RecBC heterodimers in extracts (Fig 5), although strains carrying the corresponding recC deletion alleles, recC2717 and recC2723, were recombination proficient (Fig 4). The absence of stable RecBC²⁷¹⁷ and RecBC²⁷²³ heterodimer formation may reflect inefficient assembly in cell-free extracts. We infer from these data that at least some of the amino acids in the region from 790 to 840 of RecC are important for stable assembly of RecB with RecC.

Partial proteolysis identifies two domains in RecC:
The preceding genetic results indicate that the C terminus of RecC contains a domain that is required to assemble RecD into RecBCD holoenzyme. We obtained physical evidence for such a domain by protease digestion of purified RecC polypeptide. Partial digestion of purified RecC with trypsin or subtilisin yielded two fragments of 95 and 35 kD (Fig 6, lanes 6, 7, 9, and 10). As noted previously, fragments of similar sizes were observed in purified RecC polypeptide that had been stored at 4° for several months and was cleaved as a result of trace contamination (Fig 2B, lane 3; Fig 6, lane 8). The two fragments of RecC produced by partial digestion with trypsin were physically dissociated, as indicated by analysis on a native polyacrylamide gel (data not shown).

We determined the positions of protease cleavage by N-terminal protein sequencing and mass spectroscopy. The N terminus of the 35-kD trypsin-generated fragment began with amino acid 805 (data not shown). The N terminus of the 35-kD fragment of RecC cleaved during storage began with amino acid 819 and had a molecular mass of 35,153 D (MALDI analysis, data not shown; predicted to be 35,157 D). These results locate a protease-sensitive cleavage site that corresponds to a position 22 to 35 codons from the recC1041 stop codon toward the N terminus of RecC (Fig 1). The 35-kD domain thus corresponds to the region shown above to be essential for RecD assembly into holoenzyme.

If the C terminus of RecC is required for assembly of RecD, it or full-length RecC might bind to RecD. We searched for such an interaction with the subunit assembly assay using Western blots of native polyacrylamide gels (see above and Fig 5). For these experiments we mixed extracts containing RecD polypeptide, purified RecB polypeptide, and full-length RecC polypeptide or truncated RecC polypeptides from extracts. In these assays we routinely detected RecBCD enzyme heterotrimer and RecBC heterodimer when using full-length RecC polypeptide (Fig 5B, lanes 2, 4, 6, and 8), but we failed to see any evidence of a RecC-RecD interaction in the absence of RecB polypeptide (lanes 3 and 7). Full-length RecC, the N terminus of RecC (RecC2723), or the C terminus of RecC (RecC2724) all failed to form a detectable complex with RecD (lanes 3, 7, 11, and data not shown). The almost neutral 67-kD RecD polypeptide would be expected to markedly retard the migration of the RecC polypeptide: see for example the considerable retardation of RecBCD enzyme vs. RecBC enzyme in Fig 5A. These data suggest that RecC interacts with RecD to form a heterotrimer only in the presence of RecB. We cannot exclude, however, the possibility that a RecCD complex was not detectable in this assay or that the experimental conditions were not appropriate for an interaction that occurs inside the cell. Further experiments on enzyme assembly may identify conditions that allow a RecC-RecD interaction.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have used genetic and physical assays of recBCD wild-type and recC mutant strains to identify a domain of RecC that is required to assemble the regulatory RecD subunit into RecBCD holoenzyme. This assembly into RecBCD enzyme heterotrimer is critical because the RecD subunit is required for nuclease activity, high affinity for DNA ends, and the regulation of RecBCD enzyme activities by Chi sites. Thus, the RecC-RecD interface is likely to be a major target of the RecBCD enzyme-Chi regulatory mechanism. Our work also demonstrates that the region required for assembly of RecD into holoenzyme is dispensable for recombination proficiency (Table 2); deletion of the C terminus of RecC leaves the RecBC¹⁰⁴¹ enzyme able to promote recombination through DNA unwinding and RecA loading (Fig 3). This result further defines the minimal region of RecC required for RecBC enzyme activity and reveals the role of two regions of RecC in enzyme activity.

The C-terminal region of the RecC subunit is required for enzyme assembly as demonstrated by deletion mutations in recC that remove codons 790–1122 of recC and eliminate RecD assembly into RecBCD enzyme (Fig 1, Fig 2, and Fig 5). The recC1010 point mutation in codon 905 has the same nonassembly phenotype, indicating that a single amino acid change in this region of RecC can also disrupt RecD assembly (Fig 2). RecC polypeptide contains a protease-sensitive cleavage site in the region of amino acids 805–819 (see RESULTS and Fig 1 and Fig 6). After cleavage at this site, the two fragments were physically dissociated. These results suggest that the C-terminal 35-kD portion of RecC corresponds to a separate physical domain essential for interaction with RecD. This region was dispensable for recombination proficiency, indicating that the enzyme functions of DNA unwinding and RecA loading do not require the C terminus of RecC.

The 95-kD N-terminal domain of RecC contains functions essential for recombination. The N terminus of RecC, with RecB, was sufficient to confer recombination proficiency by its DNA-unwinding and RecA-loading activities (Fig 3; Table 2). Deletions leaving fewer than 679 codons of the recC N terminus conferred a null phenotype, indicating that this region is required for some combination of DNA binding, DNA unwinding, and RecA loading (Fig 4). The N terminus of RecC is also required for Chi activity: recC* mutants, which alter codons 646–653, lack Chi genetic activity but retain nuclease activity and DNA-unwinding activity (SCHULTZ et al. 1983 ; HANDA et al. 1997 ; ARNOLD et al. 1998 , ARNOLD et al. 2000 ).

We have shown that the C-terminal domain of RecC is required for assembly of RecD into holoenzyme. We searched for, but were unable to demonstrate, a direct interaction between RecD and full-length RecC or the 35-kD C-terminal domain (Fig 5 and data not shown). Evidence for a RecC-RecD interaction was obtained, however, in a genetic study detailing the effects of overproduction of individual RecBCD enzyme subunits on Chi activity (RINKEN and WACKERNAGEL 1992 ). Chi hotspot activity in phage lambda vegetative crosses requires RecBCD holoenzyme and is detected when more recombination events occur in an interval with Chi than in the same interval without Chi (STAHL and STAHL 1977 ). RINKEN and WACKERNAGEL 1992 noted that overproduction of the RecC subunit reduced but did not eliminate Chi activity and suggested that free RecC binds RecD, thereby reducing the amount of RecD subunit available for assembly into RecBCD holoenzyme. RecBC enzyme (lacking the RecD subunit) does not act at Chi, so hotspot activity was predicted to decline, as observed (CHAUDHURY and SMITH 1984A ; AMUNDSEN et al. 1986 ; RINKEN and WACKERNAGEL 1992 ). This result supports the hypothesis that free RecC and RecD subunits interact inside the cell through, we infer, the C terminus of RecC. Our failure to detect a RecC-RecD interaction could be due to the lack of appropriate conditions used in the subunit assembly assay. In cells, partially folded RecC and RecD might interact without RecB, but only during translation. It is also possible that the RecC subunit forms a complex with RecB, onto which RecD is then assembled; when RecB was present in our assays, RecBCD holoenzyme was routinely assembled (Fig 5).

RecBCD enzyme is inactivated by interaction with Chi in an apparent two-step process that involves the RecD subunit (TAYLOR and SMITH 1992 , TAYLOR and SMITH 1999 ). The interface between the RecC and RecD subunits identified here may be a target for the Chi-dependent signals that change enzyme activity. The first step of inactivation involves a change in RecBCD enzyme activity at Chi. RecD plays a part in this step because it is required (CHAUDHURY and SMITH 1984A ; AMUNDSEN et al. 1986 ), with the C terminus of RecB, which contains at least part of the nuclease domain (YU et al. 1998A , YU et al. 1998B ; WANG et al. 2000 ), for full nuclease activity and action at Chi sites. Changes in RecBCD enzyme activity occur as the enzyme interacts with a DNA substrate containing a Chi site in the following way. RecBCD enzyme binds to the end of DNA and initiates unwinding (TAYLOR and SMITH 1980 ). During the ensuing unwinding, RecBCD enzyme recognizes Chi and makes a 3'-end for RecA loading (DIXON and KOWALCZYKOWSKI 1993 ; ANDERSON and KOWALCZYKOWSKI 1997 ) by nicking the DNA (PONTICELLI et al. 1985 ; TAYLOR et al. 1985 ) or reducing exonuclease activity (DIXON and KOWALCZYKOWSKI 1991 ). The RecD subunit inhibits RecA loading until RecBCD enzyme interacts with Chi (CHURCHILL et al. 1999 ; AMUNDSEN et al. 2000 ). The change in nuclease activity or a signal coordinated with nuclease activity appears to cause an as-yet-undetermined change in RecD that exposes a domain of RecB for RecA loading (AMUNDSEN et al. 2000 ). As proposed previously, RecD may undergo a conformational change or be released from the holoenzyme at Chi (THALER et al. 1988 ; STAHL et al. 1990 ). In either case the interface with RecC is likely to play a role in signaling this change, allowing a conformational change or disassembly of RecD. The second step of inactivation occurs after RecBCD enzyme has moved through the DNA substrate and is disassembled into its three individual subunits but only after encountering a Chi site (TAYLOR and SMITH 1999 ). The RecC-RecD interface must be disrupted at this stage of Chi-dependent inactivation of RecBCD enzyme.

Within a RecBCD enzyme molecule, separate domains may transmit signals to modulate enzyme activity. As noted above, Chi recognition changes the activity of RecBCD enzyme by as-yet-undetermined changes in the RecD subunit. The N terminus of RecC, which contains a domain for Chi recognition (ARNOLD et al. 2000 ), may transmit a signal to the C terminus of RecC, which then causes a change in RecD. Further analysis of enzyme assembly and the critical RecC-RecD interface may define more precisely how RecBCD enzyme is regulated by Chi.

	ACKNOWLEDGMENTS

We thank Roland Strong and Frederick Dahlquist for helpful discussions; Phil Gafken of the Biotechnology Center (FHCRC) and Debra McMillen of the Biotechnology Laboratory (Institute of Molecular Biology at the University of Oregon) for mass spectroscopy; Stephanie Thibodeaux for complementation analysis of recC1041; and Luther Davis, Joseph Farah, and Walter Steiner for comments on the manuscript. This work was supported by Public Health Service grant GM-31693 from the National Institute of General Medical Sciences.

Manuscript received January 9, 2002; Accepted for publication February 20, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AMUNDSEN, S. K., A. F. TAYLOR, A. M. CHAUDHURY, and G. R. SMITH, 1986  recD: the gene for an essential third subunit of exonuclease V. Proc. Natl. Acad. Sci. USA 83:5558-5562[Abstract/Free Full Text].

AMUNDSEN, S. K., A. M. NEIMAN, S. M. THIBODEAUX, and G. R. SMITH, 1990  Genetic dissection of the biochemical activities of RecBCD enzyme. Genetics 126:25-40[Abstract].

AMUNDSEN, S. K., A. F. TAYLOR, and G. R. SMITH, 2000  The RecD subunit of the Escherichia coli RecBCD enzyme inhibits RecA loading, homologous recombination, and DNA repair. Proc. Natl. Acad. Sci. USA 97:7399-7404[Abstract/Free Full Text].

ANDERSON, D. G. and S. C. KOWALCZYKOWSKI, 1997  The translocating RecBCD enzyme stimulates recombination by directing RecA protein onto ssDNA in a -regulated manner. Cell 90:77-86[Medline].

ANDERSON, D. G., J. J. CHURCHILL, and S. K. KOWALCZYKOWSKI, 1999  A single mutation RecB^D1080A, eliminates RecA protein loading but not Chi recognition by RecBCD enzyme. J. Biol. Chem. 274:27139-27144[Abstract/Free Full Text].

ARNOLD, D. A. and S. C. KOWALCZYKOWSKI, 2000  Facilitated loading of RecA protein is essential to recombination by RecBCD enzyme. J. Biol. Chem. 275:12261-12265[Abstract/Free Full Text].

ARNOLD, D. A., P. R. BIANCO, and S. C. KOWALCZYKOWSKI, 1998  The reduced levels of recognition exhibited by the RecBC¹⁰⁰⁴D enzyme reflect its recombination defect in vivo. J. Biol. Chem. 273:16476-16486[Abstract/Free Full Text].

ARNOLD, D. A., N. HANDA, I. KOBAYASHI, and S. C. KOWALCZYKOWSKI, 2000  A novel, 11 nucleotide variant of , *: one of a class of sequences defining the Escherichia coli recombination hotspot . J. Mol. Biol. 300:469-479[Medline].

BOEHMER, P. E. and P. T. EMMERSON, 1991  Escherichia coli RecBCD enzyme: inducible overproduction and reconstitution of the ATP-dependent deoxyribonuclease from purified subunits. Gene 102:1-6[Medline].

BOEHMER, P. E. and P. T. EMMERSON, 1992  The RecB subunit of the Escherichia coli RecBCD enzyme couples ATP hydrolysis to DNA unwinding. J. Biol. Chem. 267:4981-4987[Abstract/Free Full Text].

CHAUDHURY, A. M. and G. R. SMITH, 1984a  A new class of Escherichia coli recBC mutants: implications for the role of RecBC enzyme in homologous recombination. Proc. Natl. Acad. Sci. USA 81:7850-7854[Abstract/Free Full Text].

CHAUDHURY, A. M. and G. R. SMITH, 1984b  Escherichia coli recBC deletion mutants. J. Bacteriol. 118:242-249.

CHURCHILL, J. J., D. G. ANDERSON, and S. C. KOWALCZYKOWSKI, 1999  The RecBC enzyme loads RecA protein onto ssDNA asymmetrically and independently of , resulting in constitutive recombination activation. Genes Dev. 13:901-911[Abstract/Free Full Text].

DIXON, D. A. and S. C. KOWALCZYKOWSKI, 1991  Homologous pairing in vitro stimulated by the recombination hotspot, Chi. Cell 66:361-371[Medline].

DIXON, D. A. and S. C. KOWALCZYKOWSKI, 1993  The recombination hotspot is a regulatory sequence that acts by attenuating the nuclease activity of the E. coli RecBCD enzyme. Cell 73:87-96[Medline].

EICHLER, D. C. and I. R. LEHMAN, 1977  On the role of ATP in phosphodiester bond hydrolysis catalyzed by the RecBC deoxyribonuclease of Escherichia coli.. J. Biol. Chem. 252:499-503[Abstract/Free Full Text].

FINCH, P. W., A. STOREY, K. BROWN, I. D. HICKSON, and P. T. EMMERSON, 1986a  Complete nucleotide sequence of recD, the structural gene for the alpha subunit of exonuclease V of Escherichia coli.. Nucleic Acids Res. 14:8583-8594[Abstract/Free Full Text].

FINCH, P. W., R. E. WILSON, K. BROWN, I. D. HICKSON, and A. E. THOMPKINSON et al., 1986b  Complete nucleotide sequence of the Escherichia coli recC gene and of the thyA recC intergenic region. Nucleic Acids Res. 14:4437-4451[Abstract/Free Full Text].

HANDA, N., S. OHASHI, K. KUSANO, and I. KOBAYASHI, 1997  *, a -related 11-mer sequence partially active in an E. coli recC* strain. Genes Cells 2:525-536[Abstract].

HICKSON, I. D., C. N. ROBSON, K. E. ATKINSON, L. HUTTON, and P. T. EMMERSON, 1985  Reconstitution of RecBC DNase activity from purified Escherichia coli RecB and RecC proteins. J. Biol. Chem. 260:1224-1229[Abstract/Free Full Text].

HOWARD-FLANDERS, P. and L. THERIOT, 1966  Mutants of Escherichia coli K12 defective in DNA repair and genetic recombination. Genetics 53:1137-1150[Free Full Text].

JULIN, D. A. and I. R. LEHMAN, 1987  Photoaffinity labeling of the recBCD enzyme of Escherichia coli with 8-azidoadenosine 5'-triphosphate. J. Biol. Chem. 262:9044-9051[Abstract/Free Full Text].

KORANGY, F. and D. A. JULIN, 1992  A mutation in the consensus ATP-binding sequence of the RecD subunit reduces the processivity of the RecBCD enzyme from Escherichia coli.. J. Biol. Chem. 267:3088-3095[Abstract/Free Full Text].

KORANGY, F. and D. A. JULIN, 1993  Kinetics and processivity of ATP hydrolysis and DNA unwinding by the RecBC enzyme from Escherichia coli.. Biochemistry 32:4873-4880[Medline].

KOWALCZYKOWSKI, S. C., 2000  Initiation of genetic recombination and recombination-dependent replication. Trends Biochem. Sci. 25:156-165[Medline].

KOWALCZYKOWSKI, S. C., D. A. DIXON, A. K. EGGLESTON, S. D. LAUDER, and W. M. REHRAUER, 1994  Biochemistry of homologous recombination in Escherichia coli.. Microbiol. Rev. 58:401-465[Abstract/Free Full Text].

MATSUDAIRA, P., 1987  Sequence from picomole quantities of proteins electroblotted onto polyvinylidene difloride membranes. J. Biol. Chem. 262:10035-10038[Abstract/Free Full Text].

OLIVER, D. B. and E. B. GOLDBERG, 1977  Protection of parental T4 DNA from a restriction exonuclease by the product of gene 2. J. Mol. Biol. 116:877-881[Medline].

PONTICELLI, A. S., D. W. SCHULTZ, A. F. TAYLOR, and G. R. SMITH, 1985  Chi-dependent DNA strand cleavage by RecBC enzyme. Cell 41:145-151[Medline].

RINKEN, R. and W. WACKERNAGEL, 1992  Inhibition of the recBCD-dependent action of Chi recombinational hot spots in SOS-induced cells of Escherichia coli.. J. Bacteriol. 174:1172-1178[Abstract/Free Full Text].

SCHULTZ, D. W., A. F. TAYLOR, and G. R. SMITH, 1983  Escherichia coli RecBC pseudorevertants lacking Chi recombinational hotspot activity. J. Bacteriol. 155:664-680[Abstract/Free Full Text].

SMITH, G. R., 2001  Homologous recombination near and far from DNA breaks: alternative roles and contrasting views. Annu. Rev. Genet. 35:243-274[Medline].

STAHL, F. W. and M. M. STAHL, 1977  Recombination pathway specificity of Chi. Genetics 86:715-725[Abstract/Free Full Text].

STAHL, F. W., L. C. THOMASON, I. SIDDIQI, and M. M. STAHL, 1990  Further tests of a recombination model in which Chi removes the RecD subunit from the RecBCD enzyme of Escherichia coli.. Genetics 126:519-533[Abstract].

TAYLOR, A. F., 1988 RecBCD enzyme of Escherichia coli, pp. 231–263 in Genetic Recombination, edited by R. KUCHERLAPATI and G. R. SMITH. American Society for Microbiology, Washington, DC.

TAYLOR, A. and G. R. SMITH, 1980  Unwinding and rewinding of DNA by the RecBC enzyme. Cell 22:447-457[Medline].

TAYLOR, A. F. and G. R. SMITH, 1992  RecBCD enzyme is altered upon cutting DNA at a Chi recombination hotspot. Proc. Natl. Acad. Sci. USA 89:5226-5230[Abstract/Free Full Text].

TAYLOR, A. F. and G. R. SMITH, 1995  Monomeric RecBCD enzyme binds and unwinds DNA. J. Biol. Chem. 270:24451-24458[Abstract/Free Full Text].

TAYLOR, A. F. and G. R. SMITH, 1999  Regulation of homologous recombination: Chi inactivates RecBCD enzyme by disassembly of the three subunits. Genes Dev. 13:890-900[Abstract/Free Full Text].

TAYLOR, A. F., D. W. SCHULTZ, A. S. PONTICELLI, and G. R. SMITH, 1985  RecBC enzyme nicking at Chi sites during DNA unwinding: location and orientation dependence of the cutting. Cell 41:153-163[Medline].

THALER, D. S., E. SAMPSON, I. SIDDIQI, S. M. ROSENBERG, F. W. STAHL et al., 1988 A hypothesis: Chi-activation of RecBCD enzyme involves removal of the RecD subunit, pp. 413–422 in Mechanisms and Consequences of DNA Damage Processing, edited by E. FRIEDBERG and P. HANAWALT. Alan R. Liss, New York.

TOMIZAWA, J. and H. OGAWA, 1972  Structural genes of ATP-dependent deoxyribonuclease of Escherichia coli. Nature 239:14-16.

WANG, J., R. CHEN, and D. A. JULIN, 2000  A single nuclease active site of the Escherichia coli RecBCD enzyme catalyzes single-stranded DNA degradation in both directions. J. Biol. Chem. 275:507-513[Abstract/Free Full Text].

WILLETTS, N. S. and D. W. MOUNT, 1969  Genetic analysis of recombination-deficient mutants of Escherichia coli K12 carrying rec mutations cotransducible with thyA.. J. Bacteriol. 100:923-934[Abstract/Free Full Text].

WILLETTS, N. S., A. J. CLARK, and B. LOW, 1969  Genetic location of certain mutations conferring recombination deficiency in Escherichia coli.. J. Bacteriol. 97:244-249[Abstract/Free Full Text].

YU, M., J. SOUAYA, and D. A. JULIN, 1998a  The 30-kDa C-terminal domain of the RecB protein is critical for the nuclease activity, but not the helicase activity, of the RecBCD enzyme from Escherichia coli.. Proc. Natl. Acad. Sci. USA 95:981-986[Abstract/Free Full Text].

YU, M., J. SOUAYA, and D. A. JULIN, 1998b  Identification of the nuclease active site in the multifunctional RecBCD enzyme by creation of a chimeric enzyme. J. Mol. Biol. 283:797-808[Medline].

This article has been cited by other articles:

				S. K. Amundsen, A. F. Taylor, M. Reddy, and G. R. Smith Intersubunit signaling in RecBCD enzyme, a complex protein machine regulated by Chi hot spots Genes & Dev., December 15, 2007; 21(24): 3296 - 3307. [Abstract] [Full Text] [PDF]

				S. K. Amundsen and G. R. Smith Chi Hotspot Activity in Escherichia coli Without RecBCD Exonuclease Activity: Implications for the Mechanism of Recombination Genetics, January 1, 2007; 175(1): 41 - 54. [Abstract] [Full Text] [PDF]

				K. Regha, A. K. Satapathy, and M. K. Ray RecD Plays an Essential Function During Growth at Low Temperature in the Antarctic Bacterium Pseudomonas syringae Lz4W Genetics, August 1, 2005; 170(4): 1473 - 1484. [Abstract] [Full Text] [PDF]

				I. Ivancic-Bace, E. Salaj-Smic, and K. Brcic-Kostic Effects of recJ, recQ, and recFOR Mutations on Recombination in Nuclease-Deficient recB recD Double Mutants of Escherichia coli J. Bacteriol., February 15, 2005; 187(4): 1350 - 1356. [Abstract] [Full Text] [PDF]