Proc. Nati. Acad. Sci. USA Vol. 81, pp. 550-554, January 1984 Genetics
Escherichia coli mutants suppressing replication-defective mutations of the ColE1 plasmid (A-mini-ColEl hybrid phage/genetic mapping/DNA sequence/RNase H/temperature sensitivity)
SATOSHI NAITOt, TAKAKO KITANIt, TOHRU OGAWA*, TUNEKO OKAZAKIt, AND HISAO UCHIDAt§ *Institute of Medical Science, University of Tokyo, Minato-ku, Tokyo 108, Japan; and tFaculty of Science, University of Nagoya, Chikusa-ku, Nagoya 464, Japan
Communicated by Hitoshi Kihara, October 3, 1983
this communication, we describe a procedure for the isolation of host mutants that allow replication of the defective replicons, intending to identify elements recognizing the cer mutant DNA. During the survey, we found that one ColEl mutation, called cer6, was suppressed by the loss of RNase H activity from the host bacteria. In cells defective in the RNase H activity, the wild-type ColEl replicon replicated.
Mutants of Escherichia coli K-12 have been ABSTRACT isolated that suppress cer mutants, ColEl mutants that are unable to replicate as the plasmid. These host suppressors were designated her, for host factor affecting ColE1 replication. Each her suppressor showed a characteristic pattern of suppression depending on the cer mutation used for selecting the mutant bacteria. One of the suppressors, named herA, that suppressed cer6, a single-base-pair alteration 160 base pairs upstream of the ColEl replication origin, was genetically identified as an alteration of the rnh gene (RNase H). HerA was recessive to its wild-type allele. RNase H activity of herA cell extracts was defective. Conversely, rnh mutants that were isolated independently of CoIEl replication supported replication of cer6 DNA. Some rnh mutants manifested the HerA phenotype only above a certain transition temperature, and their RNase H activity was found to be temperature sensitive. Therefore, replication of cer6 DNA in vivo is sensitive to RNase H activity. Under the conditions that suppressed cer6, the wild-type colEl replicon replicated normally. Then, ColEl replication in vivo proceeds in the absence of RNase H activity, which has been shown to be required for in vitro replication of the DNA.
MATERIALS AND METHODS
mainly directed toward establishment followed by biochemical dissection of soluble systems that support DNA replication in vitro, possibilities of characterizing elements required in vivo have not been extensively explored. DNA replication of ColEl plasmid of Escherichia coli is initiated in vitro by successive actions of E. coli RNA polymerase, RNase H, and DNA polymerase I (1, 2). Based on detailed analysis of the reactions, Itoh and Tomizawa proposed the following model to describe the initial reactions of ColEl DNA replication in vitro (1, 2). Starting 555 base pairs (bp) upstream of the replication origin (the ori site) (3), RNA polymerase synthesizes a transcript that is cleaved by RNase H (4-6) and is then used as a primer by DNA polymerase I at the ori site. Genetic evidence corroborating the validity of the proposed model to describe in vivo events is as follows: (i) ColEl DNA replication in wild-type bacteria is sensitive to rifampicin but not in RNA polymerase mutants resistant to the drug. Therefore, RNA polymerase is needed for plasmid replication (7). (it) ColEl DNA cannot be maintained in polAl bacteria, indicating that DNA polymerase I is required for maintenance of the plasmid (8). On the other hand, there is no genetic evidence for the suggested RNase H requirement. We have previously reported isolation of mini-ColEl mutants that are defective in DNA replication (9). These mutations were designated cer, for ColE1 replication defective. In
Bacteria, Plasmid, and Bacteriophage Strains. The E. coli K-12 strains used and their relevant characteristics were as follows: N211(Hfr Cavalli, recAl sup0) (10); KH5401(F- thr ilv sup'), a sup' derivative of KH54 (11), was supplied by M. Imai; KL16-99(Hfr recAl) (12); KY4644(leu proA recAl/KLF4) (13). NBSS(metB metD dapD his proA supE strr) was prepared by introducing metBi and metD88 of CD4 [CGSC 5096 (14)] and dapD4 of AT982 (15) into AB1133 (16) (N. Kuwabara, personal communication). ON121 and 0N152 were prepared by introducing rnh59 proA+ and rnh9l proA+ into ON112(thi proA metE supE endA rna) (17). The cer mutants of pAO3 (18) were as described (9). The position of a particular nucleotide on plasmid pAO3 is designated by the number of nucleotides that separate it from the ori site (3), with a plus or minus sign, depending on whether the position is downstream or upstream of ori. XVIII has a single EcoRI site (srI-2) (19); a hybrid phage, XVIII-pAO3, was called XSN4 (9); Xpsu+2 carries supE suppressor (20); AdnaQ+ rnh+ (21), XdnaQ::Tn3 rnh+ (21), and XdnaQ+ rnh::Tn3 (22) were gifts from T. Horiuchi. Reagents. Sources of enzymes used were as follows: Sst I and Ava I (Bethesda Research Laboratories); Hga I (New England BioLabs); and EcoRI, BamHI, Hinfl, T4 DNA polymerase, and T4 ligase (Takara Shuzo, Kyoto, Japan). DNA oligomer d(C-G-G-A-A-T-T-C-C-G) was obtained from Genex, Rockville, MD. Bio-Rad Protein Assay was purchased from Bio-Rad. Ampicillin (Ap) and tetracycline (Tc) were obtained from Banyu Pharmaceutical, Tokyo, Japan, and Sigma, respectively. N-methyl-N'-nitro-N-nitrosoguanidine was obtained from Aldrich. Enzymes and reagents were used as recommended by the supplier. Media and buffers were used as described (9). Construction of AVIII*Ap and AVIII*Tc Vector Phages. XVIII (19) has a single EcoRI site at 54.3% the length of the X genome (54.3% X) and retains the attP sequence. Circularized XVIII DNA was first partially digested with BamHI. The ends created were filled in by T4 DNA polymerase and ligated to EcoRI linker d(C-G-G-A-A-T-T-C-C-G). The mixture was digested completely with EcoRI and ligated again. The resultant phage, named XVIII*, has a single EcoRI site at 54.3% X, but the region from 54.3% X to a BamHI site at 58.1% X is deleted (N. Inoue, personal communication). Therefore, the phage lacks attP and part of the int gene. The
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Abbreviations: bp, base pair(s); Ap, ampicillin; Tc, tetracycline; ccc, covalently closed circular. tTo whom reprint requests should be addressed.
Initiation of DNA replication is a complex but controlled process that usually starts from a unique site(s) on a genome. Although recent efforts to understand DNA replication are
Proc. NatL. Acad. Sci. USA 81 (1984)
Genetics: Naito et aL ,3lactamase gene of pBR322 (23), 1,192 bp long and flanked by EcoRI and HgaI sites, was purified and cloned into XVIII* DNA between the EcoRI site at 54.3% X and the Sst I site at 51.5% X. The phage was named XVIII*Ap. Similarly, the Tc-resistance gene of pBR322, 1,426 bp long and flanked by EcoRI and Ava I sites, was cloned into XVIII* at the same position, and the phage obtained was named XVIII*Tc. pAO3 (18) or its mutants were cloned into XVIII*Ap or XVIII*Tc at their unique EcoRI sites. The polarity of the cloned fragment was determined by examining Hinfl digests, and those clones having leftward direction for RNA I transcription (18) were used. DNA Sequence Analysis. The DNA sequence of cer mutants of pAO3 from position +75 to -662 was determined by the chemical degradation method ofMaxam and Gilbert (24). RNase H Assay. The standard reaction mixture (50 1ul) for RNase H assay was 40 mM Tris HCl, pH 7.8/10 mM MgCl2/0.5 mM dithiothreitol/20 mM NaCl/bovine serum albumin (0.1 mg/ml)/7.6 ,uM [3H]poly(C) (11.4 cpm/pmol of nucleotide) hybridized to 20 ,uM poly(dG) and 10 ,l of appropriately diluted S-100 extract (17). By definition, 1 unit of the enzyme converts 1 nmol of [3H]poly(C) to 5% trichloroacetic acid-soluble form in 15 min at 30°C. Bio-Rad Protein Assay-was used to estimate the protein concentration of the S-100 extract.
RESULTS Isolation of Bacterial Mutants Allowing Replication of pAO3 Mutant Replicons. The pAO3 is a 1,683-bp segment of the ColEl plasmid that replicates autonomously in E. coli (18). We previously described mutants of pAO3 that are defective in DNA replication (9), and we designated them cer. To select for bacterial mutants allowing replication of the cer mutant replicons, each cer DNA was cloned into X vectors harboring the Ap- or Tc-resistance gene, and the mutant bacteria were initially isolated as drug-resistant colonies on infection of lysogens with these tester phages. These host mutations were named her for host factor affecting ColE1 replication. Isogenic tester phages XVIII*Ap-pAO3cer and XVIII*Tc-pAO3cer were prepared by EcoRI digestion of XVIIIpAO3cer DNA (9) and by recloning into XVIII*Ap and XVIII*Tc vectors, respectively. E. coli strain N211, a Cavalli-type Hfr harboring recAl, was lysogenized with Xpsu+2 phage, which transduces supE suppressor, and was used as the parental strain to isolate host mutants suppressing defective replication of cer DNA. Because cI of XVIII* sustained an amber mutation, the supE suppressor was introduced to prevent possible titration of the prophage repressor by replication of superinfecting tester phages. A two-step selection procedure was used to select her mutants; in the first step, N211(Xpsu+2) cells mutagenized by treatment with N-methyl-N'-nitro-N-nitrosoguanidine were infected with XVIII*Tc-pAO3cer phages at a multiplicity of about 0.5 phage per cell. Tc-resistant transductants were selected on TY/glc agar containing Tc at 5 Mg/ml. In the second step, Tc-resistant colonies were replica-plated onto TY/glc agar containing Ap at 20 ,g/ml which had been spread with XVIII*Ap-pA03cer harboring the same cer mutation at _107 per plate. In a typical experiment with the cer6 mutation (9), -1.5 x 109 tester phages were used in the first selection, and -6,000 Tc-resistant colonies were obtained. Of these, 13 were Ap resistant in the second step. The remaining colonies may have been those harboring revertants of cer6 that were present among the first tester phages. Of 13 Ap-resistant colonies, 9 were Tc resistant. They may have been revertants of recAl among the parental bacteria that
the first and the second tester
clones were those
in which the first tester
replicating in the plasmid state was excluded by the superinfecting second tester phages because of plasmid incompatibility. The mutations isolated were designated her followed by the isolate number. Thus, the four mutations obtained were named her34, her35, her38, and her39. The cer6 mutation was a single bp alteration, C-G to TEA, at position -160 (9). Other cer mutations described previously have since been characterized. Among them, cerlO was an insertion of one APT pair within a stretch of seven consecutive A-T's from position -576 to -582; cerl7 was a deletion of one AFT from the same APT stretch. No other alteration was found for both cerlO and cerl7, as well as for cer6, between an Ava II site at position -662 and an Hinfl site at position +74. [3H]Thymine incorporation of cells infected in the presence of chloramphenicol by the cer phages showed that DNA synthesis driven by the plasmid replicon is very much decreased by cer6, cerlO, or cerl7 mutations (data not shown). Host suppressors for cerlO and cerl7 were isolated as described above, and they were named her8l, her85, her87, and her89 for cerlO; her62 and her63 for cerl7. Each her mutant was purified by single colony isolation and cured of the tester phages used in its isolation. Specificity of her Mutations. Each her mutant was tested for ift-ability to suppress various cer mutations by measuring transduction frequency of the Ap resistance on infection with XVIII*Ap-pAO3cer phages. The results (Table 1) showed that every her mutant selected using cer6 phage failed to be transduced by tester phages harboring a cerlO or cerl7 mutation. On the other hand, those her mutants that were selected by suppression of ceri7 were not transduced to Ap resistance by tester phages harboring either a cer6 or cerlO mutation. The her mutants that were selected by suppression of cerlO were able to suppress both cerlO and ceri7 but not cer6. The tester phage having the wild-type (cer+) pAO3 transduced Ap resistance with high frequencies to every her mutant; the vector phage XVIII*Ap that does not contain the pAO3 fragment transduced the resistance to none of the mutant strains (data not shown). Therefore, it appears that each her suppressor shows a characteristic pattern of suppression depending on the cer mutation used for its selection. The properties of her suppressors for the cer6 mutation will be described in more detail. Suppression of DNA Replication of pAO3cer6 Mutant Replicon. Bacteria were infected with XVIII-pAO3 or XVIIIpAO3cer6 phages in the presence of chloramphenicol, and Table 1. Frequency of Ap-resistance transduction of her mutant bacteria by various cer tester phages Bacteria Tester phage Selectiont Mutant cer6 cerlO cerl 7 cer+ None her+ 8 x 10-7 7 x 10-5 2 x 10-5 0.36 cer6 her34 0.30 3 x 10-5 2 x 10-5 0.14 her35 0.39 3 x 10-5 1 x 10-5 0.29 her38 0.22 4 x 10-5 2 x 10-5 0.14 her39 0.19 6 x 10-5 3 x 10-5 0.17 cerlO her8l 6 x 10-6 0.007 0.034 0.66 her85 4 x 10-6 0.084 0.24 0.60 cerl7 her62 3 x 10-6 7 x 10-5 0.32 0.37 her63 4 x 10-7 9 x 10-5 0.20 0.47 XVIII*Ap-pAO3 tester phages harboring various cer mutations were added at a multiplicity of 0.05 per cell to bacterial strains that were suspended in TM buffer at 1 x 109 cells per ml, and adsorption was allowed to continue for 20 min at 370C. Samples were plated on TY/glc agar containing Ap at 20 ,g/ml and incubated at 370C. To detect suppression of the cerlO mutation, Ap at 10 pg/ml was added to agar. Number of resistants divided by number of tester phages used was designated as the frequency of transduction. tThe cer allele of tester phages used for selecting her mutant bacteria.
Genetics: Naito et aL
Proc. NatL Acad Sci. USA 81 (1984) Table 2. Dominance test for herA39 mutation
Frequency of transductiont herA allele 0.21 herA39 9 X 10-6 herA39/herA' SN59(X)/KLF4 SN59(X) F- was prepared by P1 cotransduction of herA39 with metD' into NB55(X), and then introducing recAl by conjugation with KL16-99. SN59(X)/KLF4 was prepared by sexducing SN59(X) with KY4644 and selecting for Dap+ merozygotes. tPhage used was XVIII*Ap-pA03cer6, and frequency of transduction of Ap resistance was measured as in Table 1. Bacteria SN59(X) F-
FIG. 1. DNA synthesis in the presence of chloramphenicol. Thymine-requiring cells were grown at 37°C to 5 x 108 cells per ml in Casamino acids medium supplemented with thymine (5 ,ug/ml)/diaminopimelic acid (40 Ag/ml)/0.2% maltose. Twenty minutes after addition of chloramphenicol (250 ug/ml), cells were collected by centrifugation and suspended in one-half the original volume of TM buffer (0.2 ml) containing chloramphenicol (250 ,g/ml). Phages were added at a multiplicity of 6 phages per cell and were allowed to adsorb for 20 min at 37°C. Phage-infected cells were then diluted 1:10 with prewarmed Casamino acids medium supplemented with thymine (5,ug/ml)/[3Hlthymine (5 ,uCi/ml; 1 Ci = 37 GBq)/diaminopimelic acid (40 ,g/ml)/0.1% glucose/chloramphenicol (250 jug/ml). After aeration at 37°C for 3 hr, cells were collected by centrifugation and DNA was extracted as described (9) and was fractionated by CsCl/ethidium bromide equilibrium centrifugation of the total cell lysates. Phages used were XVIII-pAO3 (A) and XVIIIpAO3cer6 (B). Bacteria used were SN65(A)thyA her+ (e) and SN66(A)thyA herA39 (o). They were isogenic thymine-requiring derivatives of SN50(A) and SN51(A), respectively (see legend to Table 3). Position of ccc molecules is indicated in the figure.
DNA synthesized after infection was labeled with [3H]thymine. Isogenic bacteria harboring the her+ allele were used as a control. After incubation for 3 hr, DNA was extracted and fractionated by CsCl/ethidium bromide equilibrium centrifugation. If infecting phage DNA replicates in a plasmid state, radioactivity will be found at the position expected for the covalently closed circular (ccc) DNA. As shown in Fig. 1, for the sample from her+ cells infected with XVIIIpAO3cer6 phage, essentially no radioactivity was detected in fractions corresponding to ccc molecules. Substantial radioactivity, however, was detected in the fractions in which her39 cells were used. The wild-type replicon XVIII-pAO3 replicated in a plasmid state in both her+ and her39 cells almost to the same extent. These results indicated that the her39 mutation suppressed defective DNA replication of the dnaQ, rnh 5
daw:)D 6.2 2 1 21
6 p roA
<0.3 0.3 FIG. 2. Transductional mapping of the herA39 mutation with phage P1. A selected marker is shown at the tail and an unselected marker at the head of an arrow. The number adjacent to an arrow represents the percent cotransduction. The scale, calibrated in minutes, shows a part of the 100-min standard map of E. coli K-12 (25).
cer6 mutant replicon and that the suppressor had apparently no effect on replication of the wild-type pAO3. P1 Mapping and Recessiveness of the her39 Mutation. By conjugation, the her39 suppressor isolated on a Cavalli-type Hfr strain [N211(Xpsu+2)] was transferred to a F- recipient [KH5401(Xpsu+2)] as an early marker (data not shown). One of the transconjugants that received her39 was used as the donor strain in a P1 transduction experiment using NB55(X) as a recipient. The results (Fig. 2) indicate that the suppressor is located on the standard map of E. coli (25) at 5 min between metD and proA. Hereafter, we refer to the suppressor gene as herA. The locus thus determined is in the region of two closely linked genes, dnaQ (14) and rnh (5). Other mutations that suppressed cer6 mapped in the same metDproA region by P1 transduction. On the other hand, suppressors for cerlO and cerl7 were not located within this region. A merozygote heterozygous for the herA allele was made by sexducing SN59(A) F- herA39 dapD recAl with KLF4 that has the chromosomal fragment between thr and proB (12) and selecting for Dap+. The suppression activity of the herA39 disappeared by introducing the herA+ allele (Table 2). When the cell was cured of KLF4 by acridine orange treatment (26), the suppression activity reappeared. Therefore, the herA39 mutation is recessive to its wild-type allele. Identification of the herA Gene as the Gene for RNase H (rnh). Precise mapping of the herA gene was possible because the two closely linked genes dnaQ and rnh have been isolated as a 1.5-kilobase DNA fragment and cloned into a X phage vector to give a transducing phage coding for two extra proteins of Mr 25,000 and 21,000, corresponding to products of dnaQ and rnh, respectively (21). Moreover, derivatives of the transducing phage in which each of the two genes was inactivated by insertion of transposon Tn3 (27) have been prepared (T. Horiuchi, personal communication; ref. 21). Complementation experiments carried out by lysogenization of various transducing phages to the herA39 bacteria showed that herA is an alteration of the rnh gene (Table 3). Other mutants, her34, her35, and her38, were examined in the same manner, and they were also found to be mutations of the rnh gene. Therefore, all the mutant bacteria allowing replication of the cer6 mutant replicon were altered in RNase H. HerA Is RNase H-Defective and Vice Versa. RNase H activity of the herA39 mutant cells was assayed and was found to Table 3. Complementation tests of the herA39 mutation Frequency of transductiont Bacteria 2.3 x 10-3 SN50(A) 0.51 SN51(A) 5.8 x 10-3 SN51(X)(XdnaQ+ rnh+) 6.6 x 10-3 SN51(X)(XdnaQ::Tn3 rnh+) 0.57 SN51(A)(XdnaQ+ rnh::Tn3) SN51(X) was prepared by P1 cotransduction of metD+ and herA39 into NB55(X). SN50(A) was an isogenic metD+ herA+ transductant.
used was XVIII*Tc-pAO3cer6, and Tc-resistant transductants were selected.
Genetics: Naito et aL Table 4. RNase H activity and suppression of cer6 replication RNase H activityt Frequency of Mutation units per mg of protein Bacteria transductiont herA+ SN50 23.5 (100%) 7 x 10-5 SN51 herA39 0.05 (0.2%) 0.74 rnh' 19.7 (100%) 3 x 10-4 ON112 5 x 10-4 rnh59 0.08 (0.4%) ON121 ND (<0.02%) rnh9l ON152 0.75 ND, not detectable. tCells were grown at 30'C, and RNase H activity was measured at 300C. tCells were lysogenized with X, and frequency of transduction of Ap resistance by XVIII*Ap-pAO3cer6 was measured at 370C.
be decreased as compared to that of the parental cells (Table 4). Therefore, the HerA phenotype of rnh mutants isolated independently of the ColEl replication was examined. The strain named ON152 (rnh91), which has an undetectable level of RNase H activity, was a typical HerA mutant. On the other hand, ON121 (rnh59), which showed trace RNase H activity (0.08 unit per mg of protein) could not support replication of the cer6 replicon. Because the bacteria used for the enzyme assay were grown and assayed at 30°C, whereas transduction by XVIII*Ap-pAO3cer6 was carried out at 37°C, possible thermodependency of rnh59 and herA39 mutations was tested by using various incubation temperatures. We may infer from the results summarized in Fig. 3 that the mutant cells manifest the HerA phenotype above a certain transition temperature, that the transition temperature is higher for rnh59 than that for herA39, and that the cer+ replicon replicated in both mutant strains within the temperature range tested. It is intriguing to observe that the efficiency of transduction was approaching a plateau value as the temperature increased, and it did not decline. The HerA phenotype of ON152 and ON121 disappeared when the bacteria were lysogenized with XdnaQ::Tn3 rnh+ transducing phage but not when lysogenized with XdnaQ+ rnh::Tn3 (data not shown). RNase H activity of ON121 cells grown at 37°C or at 43°C was assayed at 30°C, and found to be 0.03 units per mg of protein and less than the limit of detection (<0.002 units per mg of protein), respectively. Therefore, at least production of RNase H by ON121 was thermosensitive. The property is consistent with the observation that the herA mutation was recessive to its wild-type allele.
DISCUSSION To correlate functional steps in the initiation of DNA replication with the DNA sequence, we isolated her mutants of E. coli that suppress replication defects of the cer mutations in the ori region of the ColEl plasmid (9). Although only a few her mutants were isolated and one group of them, herA, has been characterized so far, they proved the present approach to be versatile in the study of ColEl replication in vivo. First, each her suppressor showed a characteristic pattern of suppression depending on the cer mutation used for selecting the mutant bacteria (Table 1), suggesting that the DNA sequence altered by each cer mutation has a distinct function. Two mutations, cerlO and cerl7, occurred at the same locus on the ColE1 DNA; the former was an insertion and the latter was a deletion of one A-T pair within 7 consecutive A-T's located from position -576 to -582. The region separates the -10 and the -35 regions of the promoter for synthesizing the 555-nucleotide-long primer RNA that has been characterized by in vitro analysis (2). It is interesting to note that the her mutants selected by suppression of cerlO suppressed both cerlO and cerl7, whereas the her suppressors isolated by using ceri 7 suppressed only ceri 7. On the
Proc. Natl. Acad. Sci. USA 81 (1984)
other hand, they could not suppress cer6, which is a single bp alteration at position -160, and suppressors for cer6 could not suppress cerlO and cerl7 (Table 1). Second, herA, which suppressed cer6, occurred in the rnh gene of E. coli, and extracts of herA cells were defective in RNase H activity. Conversely, mutants defective in RNase H activity that were isolated independently of ColEl replication supported replication of cer6 DNA (Table 4; Fig. 3). Under the conditions that suppressed cer6, the wild-type ColEl replicon replicated normally. RNase H is absolutely required for in vitro replication of the DNA using purified enzymes (1, 2). Although the specificity of the RNase H assay was very much increased by using [3H]poly(C)-poly(dG) as substrate (17), we could not discriminate trace activity from no activity. On the other hand, Itoh and Tomizawa (1) observed that addition of excessive amounts of RNase H to their in vitro system was detrimental for correct initiation of the wild-type ColEl replication. Therefore, it may be argued that the presence of low but positive activity of RNase H was still required for replication of the cer6 DNA that became supersensitive to the enzyme activity. The recessiveness of the herA mutation to its wild-type allele (Table 2) is consistent with this interpretation. For thermodependent herA mutants, we may consider that the intracellular activity of RNase H decreased monotonically as the incubation temperature increased, and at the lowest temperature manifesting the HerA phenotype, the permissive level of the intracellular enzyme was reached. If this permissive level were low but positive, we may expect an optimal range of incubation temperature as the temperature was further increased. On the contrary, the results for herA39 cells (Fig. 3) showed that
3'0 ...35 3'7 4'0 4.3 Temperature, IC
FIG. 3. Temperature dependency of the HerA phenotype. Cells grown at 37°C were infected with tester phages at a multiplicity of -0.05 per cell, and adsorption was allowed to continue for 20 min at 37°C. The infected cells were plated on TY/glc agar containing Ap at 20 ,ug/ml and the number of Ap-resistant colonies was scored after incubation overnight at various temperatures. Transduction frequencies were calculated as in Table 1. Bacteria and tester phages used were SN51(X)dap+ infected with XVIII*Ap-pAO3cer+ (o), SN51(X)dap+ infected with XVIII*Ap-pAO3cer6 (e); ON121(X) infected with XVIII*Ap-pAO3cer+ (A), and ON121(X) infected with XVIII*Ap-pAO3cer6 (A).
Genetics: Naito et aL
once the temperature exceeded a transition point, the HerA phenotype continued to be expressed maximally. The easiest interpretation of the results is to postulate that ColEl replication in vivo proceeds in the absence of RNase H activity. It was clear that ColEl replicon replicated in herA cells as plasmids. As a step in our isolation of her mutants, resident cer DNA was eliminated by superinfection of the cell with tester phages having the same cer DNA and resistance to different antibiotics. As shown in Fig. 1, cer DNA cloned into a X replicon replicated in herA as ccc molecules in the presence of chloramphenicol. Replication of ColEl DNA has been known to be insensitive to inhibition of protein synthesis in bacteria (28), while X is not. Furthermore, three ColEl-type plasmids [pBR322 (23), pML21 (29), and pAO43 (18)] were stably maintained in ON152 cells and their copy number was the same as in ON112 (wild-type) cells (17). Preliminary experiments (unpublished results) indicated that ColEl replication in the wild type as well as mutants defective in RNase H started from the ori site; the possibility of initiation from another site was not obvious. Hillenbrand and Staudenbauer (30) recently reported an in vitro study suggesting that RNase H acts as a discrimination factor inhibiting initiation events occurring at sites other than ori. According to their model, the primer RNA is protected from degradation by RNase H by forming an extensive secondary structure of the template-primer complex at the ori region. Although replication of our cer6 DNA was sensitive to RNase H activity in vivo, its DNA sequence near ori was not altered. Therefore, the sensitivity of cer6 is difficult to explain by this model. Because the cer6 mutation could be considered to destabilize a possible hairpin structure that starts by hydrogen bonding of bases at positions -160 and -187 (9), the mutation may have introduced a new site sensitive to RNase H by exposing sequences of the primer RNA to become available for hybridization with DNA. We thank T. Horiuchi for providing us with transducing phages and information prior to publication. Thanks are also due to N. R. Cozzarelli and B. M. Alberts for advice on preparation of the manuscript. This work was supported in part by Grants-in-Aid for Special Research Project from the Ministry of Education, Science and Culture of Japan. 1. Itoh, T. & Tomizawa, J. (1979) Cold Spring Harbor Symp. Quant. Biol. 43, 409-417. 2. Itoh, T. & Tomizawa, J. (1980) Proc. Natl. Acad. Sci. USA 77, 2450-2454.
Proc. NaiL Acad Scd USA 81 (1984) 3. Tomizawa, J., Ohmori, H. & Bird, R. E. (1977) Proc. Natl. Acad. Sci. USA 74, 1865-1869. 4. Berkower, I., Leis, J. & Hurwitz, J. (1973) J. Biol. Chem. 248, 5914-5921. 5. Carl, P. L., Bloom, L. & Crouch, R. J. (1980) J. Bacteriol. 144, 28-35. 6. Darlix, J. (1975) Eur. J. Biochem. 51, 369-376. 7. Clewell, D. B. & Evenchik, B. (1973) J. Mol. Biol. 75, 503513. 8. Kingsbury, D. T. & Helinski, D. R. (1970) Biochem. Biophys. Res. Commun. 41, 1538-1544. 9. Naito, S. & Uchida, H. (1980) Proc. Natl. Acad. Sci. USA 77, 6744 6748. 10. Miura, A. & Tomizawa, J. (1968) Mol. Gen. Genet. 103, 1-10. 11. Horiuchi, T. & Nagata, T. (1973) Mol. Gen. Genet. 123, 89110. 12. Low, K. B. (1968) Proc. Natl. Acad. Sci. USA 60, 160-167. 13. Wada, C., Hiraga, S. & Yura, T. (1976) J. Mol. Biol. 108, 2541. 14. Horiuchi, T., Maki, M. & Sekiguchi, M. (1978) Mol. Gen. Genet. 163, 277-283. 15. Shizuya, H., Livingston, D. M. & Richardson, C. C. (1974) Proc. Natl. Acad. Sci. USA 71, 2614-2617. 16. Bachmann, B. J. (1972) Bacteriol. Rev. 36, 525-557. 17. Ogawa, T. & Okazaki, T. (1983) Mol. Gen. Genet., in press. 18. Oka, A., Nomura, N., Morita, M., Sugisaki, H., Sugimoto, K. & Takanami, M. (1979) Mol. Gen. Genet. 172, 151-159. 19. Murray, N. E. & Murray, K. (1974) Nature (London) 251, 476-481. 20. Inokuchi, H., Yamao, F., Sakano, H. & Ozeki, H. (1979) J. Mol. Biol. 132, 649-662. 21. Horiuchi, T., Maki, H., Maruyama, M. & Sekiguchi, M. (1981) Proc. Natl. Acad. Sci. USA 78, 3770-3774. 22. Maki, H., Horiuchi, T. & Sekiguchi, M. (1983) Proc. Natl. Acad. Sci. USA 80, 7137-7141. 23. Sutcliffe, J. G. (1979) Cold Spring Harbor Symp. Quant. Biol. 43, 77-90. 24. Maxam, A. M. & Gilbert, W. (1980) Methods Enzymol. 65, 499-560. 25. Bachmann, B. J. & Low, K. B. (1980) Microbiol. Rev. 44, 156. 26. Hirota, Y. (1960) Proc. Natl. Acad. Sci. USA 46, 57-64. 27. Heifron, F., McCarthy, B. J., Ohtsubo, H. & Ohtsubo, E. (1979) Cell 18, 1153-1163. 28. Blair, D. G., Sherratt, D. J., Clewell, D. B. & Helinski, D. R. (1972) Proc. Natl. Acad. Sci. USA 69, 2518-2522. 29. Hersfield, V., Boyer, H. W., Chow, L. & Helinski, D. R. (1976) J. Bacteriol. 126, 447-453. 30. Hillenbrand, G. & Staudenbauer, W. L. (1982) Nucleic Acids Res. 10, 833-853.