Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
DNA replication wikipedia , lookup
DNA repair protein XRCC4 wikipedia , lookup
DNA polymerase wikipedia , lookup
DNA profiling wikipedia , lookup
Homologous recombination wikipedia , lookup
DNA nanotechnology wikipedia , lookup
Zinc finger nuclease wikipedia , lookup
United Kingdom National DNA Database wikipedia , lookup
Journal of General Microbiology (1983), 129, 1497-1 5 12. Printed in Great Britain 1497 The Mechanism of Insertion of a Segment of Heterologous DNA into the Chromosome of Bacillus subtilis By M I C H A E L Y O U N G Department of Botany and Microbiology, University College of Wales, Aberystwyth SY23 3DA, U.K. (Received 25 August 1982; revised 20 October 1982) An Escherichia coli plasmid, p1949, that is derived from pMB9 and pC194, and unable to replicate in Bacillus subtilis, can give rise to stable CmRtransformants of the latter species if it is inserted into the bacterial chromosome. A purified segment of the B. subtilis chromosome, with transforming activity against pheAl and nic-38 recipients, was used to direct the insertion of p1949 into the B. subtilis chromosome. Insertion of the ligated DNA segments occurred in the region of the chromosome from which the purified phe-nic segment was derived. Many of the properties of the resulting CmRtransformants of B . subtilis are consistent with the occurrence of a Campbell recombination mechanism leading to integration. However, certain of these properties are more easily explained if it is proposed that integration occurs by a reciprocal recombination event involving a linear ligation product. Evidence is presented suggesting that the inserted sequences may be tandemly duplicated. This may effectively vitiate the use of p1949 as a convenient means for complementation analysis of recessive mutations in B. subtilis. INTRODUCTION Several investigators have constructed plasmids that contain an antibiotic resistance determinant that is expressed in a particular host even though the plasmid is unable to replicate in that host (e.g. Haldenwang et al., 1980; Niaudet & Ehrlich, 1979). Plasmids of this type may be termed insertion vectors, since stable antibiotic-resistant progeny can be obtained in the nonpermissive host if the plasmid integrates into a resident replicon. The ligation of host DNA sequences into the heterologous plasmid apparently directs insertion into the homologous region of the resident replicon. Insertion vectors have found three important uses to date. Firstly, they can be used to clone the replication origins of other replicons (e.g. see Messer et al., 1978; Niaudet & Ehrlich, 1979; Chan & Tye, 1980; Anderson et al., 1982). Secondly, they have been employed to determine the map position of cloned fragments of unknown origin (Haldenwang et al., 1980; Wilson et al., 1981). Thirdly, an insertion vector has been used to clone one DNA sequence repeatedly from different strains of the same organism (Mkjean et al., 1981). This represents an important aid to the identification of gene products encoded by cloned DNA segments. If a family of cloned fragments can be obtained in this way, each one carrying a different allelic form of the gene of interest (ideally, suppressible alleles should be cloned), then the gene product can be identified quite readily, since high-efficiency temperature-sensitive amber suppressor strains of Escherichia coli have now been isolated (Oeschger et al., 1980). There have been several reports of the integration of heterologous DNA into the Bacillus subtilis chromosome (Harris-Warwick & Lederberg, 1978; Duncan et al., 1978; Rapoport et al., 1979; Yoneda et al., 1979; Kawamura et al., 1979; Haldenwang et al., 1980; Galizzi et al., 1981). It is widely accepted that integration cannot occur in Rec- (recE-4)hosts (Haldenwang et al., 1980; Wilson et al., 1981, Niaudet et al., 1982) and that a region of sequence homology is usually necessary to direct insertion of the foreign DNA (however, see Rubin et al., 1980). However, comparatively little is known about the mechanism by which the heterologous DNA inserts into Downloaded from www.microbiologyresearch.org by 0022-1287/83/0001-0770$02.00 0 1983 SGM IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 1498 M . YOUNG the B. suhtifis chromosome. It has generally been assumed that integration is by a Campbell mechanism, by analogy with the mechanism of integration of bacteriophage ;1 into the chromosome of E. cofi (Campbell, 1962). In some cases evidence has been adduced, principally from Southern hybridization experiments, to support this contention (Haldenwang et al., 1980; Galizzi et af., 1981; Niaudet et al., 1982). The mechanism by which foreign DNA inserts into the chromosome of B. subtifis might be expected to depend on the precise physical nature of the transforming DNA. If it is provided as covalently closed plasmid DNA, then a Campbell mechanism would seem intuitively to be the most likely method of integration. However, if the source of DNA is simply a ligated mixture of vector and host fragments, then transformation by linear ligation products becomes a distinct possibility. Following on from this, if dissimilar host fragments are ligated to the ends of the vector DNA, then integration by a reciprocal recombination event will lead to deletion of the DNA segment that lies between the two host fragments. Indeed, Niaudet et a f . (1982) have reported the isolation of an ifvA auxotroph of B. subtilis that apparently arose by this kind of recombination mechanism, In this investigation further evidence has been obtained concerning the mechanism of insertion of foreign DNA into the B. subtifis chromosome. The plasmid used was p1949 (Haldenwang et af., 1980) which encodes a chloramphenicol acetyltransferase that is expressed in B . subtifis. This gerx originated from pC194 (Iordhescu, 1975); the complete nucleotide sequence of pC194 has recently been published (Horinouchi & Weisblum, 1982). The plasmid p1949 is unable to replicate in B. subtifis,whereas in E. cofi it replicates and confers resistance to both chloramphenicol and tetracycline (the latter resistance is derived from the E. cofi plasmid pMB9). A purified fragment of the B. subtifis chromosome with transforming activity in pheA2 recipients (kindly donated by J. Szulmajster and C. Bonamy) was used to direct insertion of the plasmid into the B. subtifis chromosome. This enabled direct examination of whether integration did indeed occur at the site where natural homology existed. Also, a suitably marked recipient strain was used so that genetic evidence concerning the mechanism of plasmid integration could be obtained. Many of the results are consistent with the occurrence of a Campbell integration mechanism. However, there are certain anomalies that can be explained if integration occurred via the formation of bridge structures involving linear ligation products. METHODS Strains. The strains of E. coli and B. subtilis used in this investigation are listed in Table 1. Plasmid p1949 was harboured by E. coli strain MM294, which was kindly supplied by A. L. Sonenshein. Plasmid pSC5 was constructed by cloning an EcoRI-BamHI fragment of B. subtilis DNA that had transforming activity against phrAl recipients in the plasmid vector pBR322. Plasmid pSC5 was extracted from the rl- mk- recBC derivative of E. coli C600 listed in Table 1 and was a gift from C. Bonamy. Trunsjormution. Strains of E. coli were grown to mid-exponential phase in Luria broth and then permeabilized by calcium chloride treatment followed by heat shock (Mandel & Higa, 1970). Transformed bacteria were incubated in Luria broth at 37 "C for 90 min before plating out on selective medium. The method of Anagnostopoulos & Spizizen (1961) was used for transformation of strains of B . suhrilzs. Unless otherwise stated, DNA was added at a non-saturating concentration. In the crosses in Table 6 a saturating concentration of DNA was used and in one of them (cross 1) the spontaneous transformation technique of EphratiElizur (1968) was employed, In this case the donor strain was grown overnight in Penassay broth (Difco) and 0-1 ml of the stationary phase culture was added to 1 ml of the competent recipient bacteria. This procedure resulted in a far higher frequency of congression of unselected donor markers than is normally found using a saturating concentration of purified donor DNA. It could only be used in one of the crosses in Table 6 since it depends on the application of counterselection against growth of the donor strain (this was achieved in cross 1 by the addition of 5 pg chloramphenicol ml-' to the selective medium). For the selection of CmRtransformants of E. toli the selective medium used was Luria broth containing 15 pg chloramphenicol ml-I . CmRtransformants of B. subtilis were selected on either nutrient agar supplemented with salts (the amounts added, per litre, were: FeC1,.6H,O, 0.98 mg; MgCl, .6H,O, 8.3 mg; MnClz . 4 H z 0 , 19.8 mg; NH,Cl, 535 mg; Na2S04, 106 mg; KH2P0,, 68 mg; NH,NO,, 96.5 mg) or an appropriately supplemented glucose glutsmate minimal medium [this was based on the medium of Donellan eral. (1964) as described by Piggot (1973), except that sodium lactate was replaced by glucose which was added to a final concentration of 0.2% (wiv)]. In both cases chloramphenicol was added at a concentration of 5 pg ml-I. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 Integration of heterologous DNA in B. subtilis 1499 Table 1. Bacterial strains Genotype Strain Origin/Reference E. coli endo1 thy hsdR (p1949 CmR TcR) A. L. Sonenshein M M294 C600 r- m-* thr-1 leu-6 thi-1 supE44 lacy1 P. Kourilsky via J . Szulmajster tonA2I recBC r,- m,B. subtilis 168 MB21 MB26 MB5 1 JH648 GSY225 GSY860 SAlOl S.4 102 SA 103 trpC2 metC3 leu-8 tal-I leu-8 pheAI2 rif-2 tal-1 pheAl2 nic-38 rif-2 tal-I trpC2 pheA I spoOB136 pheAl trpC2 argAl I nic-38 trpC2 spoOB136 nic-38 rif-2 tal-I trpC2 pheAl nic-38 r i f 2 CLIOl-CL110 trpC2 spoOB136 cat CL104.2 trpC2 cat Laboratory wild-type Hranueli et al. (1974) Grant (1974) D. Hranueli; MB26 transformed to Leu+ Nicwith GSY860 J . Hoch C. Anagnostopoulos C. Anagnostopoulos JH648 transformed to Phe+ with 168 D N A MB51 transformed to Phe+ with 168 D N A GSY225 transformed to R I P Nic- with SA102 D N A pSC5 used to Ligation products of p1949 transform JH648 to CmR - all strains also Phe+ (see text) Spo+ revertant of CL104 * This strain is a recBC r,- + mk-derivative of strain C600. Transduction. PBS 1 transducing lysates were prepared essentially as described by Karamata & Gross ( I 970) and used for transduction as described by Takahashi (1963). Preparation of' DNA. Plasmid DNA was extracted from strains of E . coli by preparing cleared lysates as described by Clewell & Helinski (1969), and then subjecting them to caesium chloride density gradient centrifugation in the presence of ethidium bromide. Although the plasmid used here ( ~ 1 9 4 9 )encodes a chloramphenicol acetyltransferase, a chloramphenicol (250 pg ml-' ) amplification step was routinely employed (see Niaudet & Ehrlich, 1979). DNA was isolated from strains of B. subtilis using the method described by Marmur (1961). The CL104 DNA used in the Southern hybridization experiment (Fig. 2) was extracted and purified by caesium chloride density gradient centrifugation essentially as described by Harris-Warwick et al. (1975). Restriction endonucleases. Restriction endonucleases were purchased from BRL and were used as specified by the manufacturer. For digestion of plasmid and chromosome DNA, the enzymes were used in two- and fivefold excess, respectively. Incubations were terminated by heating at 65 "C for 5 min. Ligation qf' DNA ,fragments. DNA fragments generated by restriction endonucleases were ligated at a final concentration of 22 pg DNA ml-' . This relatively low DNA concentration was chosen to promote the generation of circular products (Dugaiczyk et al., 1975). The ligation mixture contained 0.5 units T, D N A ligase (Boehringer) and 2.5 pg DNA in 114 p1 ligation buffer (66 mM-Tris/HCl, pH 7.6; 6.6 mM-MgC1,; 10 mM-DTT; 1 mM-ATP). Ligation was carried out for 20 h at 4 "C. Agnrose gel electrophoresis. DNA restriction fragments were separated according to their molecular weights by electrophoresis through 0.6% (w/v) agarose (Sigma) in a horizontal gel electrophoresis apparatus. Electrophoresis was carried out at between 1.5 and 3 V cm-I , using a buffer containing Tris (89 mM), EDTA (2.5 mM) and borate (89 mM). pH 8,3-8-5, and afterwards the gels were stained in electrophoresis buffer containing ethidium bromide (0.5 pg ml-1). Stained gels were photographed using a Polaroid camera mounted above a UV transilluminator (UV Products, San Gabriel, CA, U.S.A.). Nii-k-translation. DNA samples were radioactively labelled to high specific activity [about 5 x 10' d.p.m. (pg DNA)-'] with 3?P-labelled ATP (Amersham) using the procedure described by Jeffries & Flavell (1977). Southern hj&ridi:ution. DNA-DNA hybridization experiments were carried out as follows. D N A restriction fragments (about 1 pg chromosome DNA and about 0.25 pg p1949 or ADNA) were separated by agarose gel electrophoresis and then depurinated in siru as described by Wahl et al. (1979). The DNA was then transferred to a nitrocellulose membrane (Schleicher & Schull) using the method of Southern (1 975). The filter was cut into suitabIe strips and the procedures used for hybridization and washing were essentially as described by Sager et al. (19811, except that dextran sulphate was omitted from the hybridization solution. The filter strips were then reassembled and autoradiographed at - 20 "C by exposure to pre-fogged Kodak X-omat R P film using an intensifying screen. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 1500 M . YOUNG E S 10.201 2.0 H Fig. 1. Restriction maps of p1949 and pSC5. The approximate positions of restriction enzyme cleavage sites were determined from the apparent molecular weights of DNA fragments generated by single and double digestion experiments; they do not necessarily correspond exactly to the precise positions determined by nucleotide sequence analysis (where known). The exact position of the CmRdeterminant of p1949 is not known; it has been assigned to its most probable position from the limited information available (Horinouchi & Weisblum, 1982; Haldenwang et af., 1980; see results). Positions have been assigned to the pheA and nic genes on pSC5, in accordance with the data in Table 3. The numbers represent MDal. B, BarnHI; E, EcoRI; H, HindIII; P, PstI; and S, SafGI. RESULTS Characterization of plasmids p1949 and pSC.5 Haldenwang et al. (1980) constructed the plasmid p1949 in three stages, the net result being the union of pMB9 (TcR, 3.55 MDal) that had been linearized with HpaI, and pC194 (CmR, 2-0 MDal) that had been linearized with HpaII. As expected, the resultant chimaera is unable to transform B. subtilis to CmR(see Table 4), since the HpaII (= MspI) recognition sequence lies in that region of pC194 that is thought to contain the origin of replication (Horinouchi & Weisblum, 1982). However, p1949 has two unexpected features: it is somewhat smaller than expected from the sizes of the two parental molecules; and it has an additional site for BamHI that is not present in either of the parent plasmids (Fig. 1). A 3-2 MDal EcoRI-BamHI fragment of B . subtilis DNA was used to direct the insertion of p1949 sequences into the B. subtilis chromosome. This fragment was cloned into pBR322 and the resultant plasmid, pSC5, was shown to have transforming activity against apheAI recipient of B. subtilis (J. Szulmajster & C. Bonamy, personal communication). This finding is confirmed in Table 2, which also shows that pSC5 will generate Nic+ transformants with a nic-38 recipient strain; however, no Phe+ transformants were obtained with a pheAl2 recipient strain. This suggests that the B. subtilis chromosomal fragment present in pSC5 does not carry an intact copy of the pheA gene. The plasmid failed to generate Spo+ transformants with the spoOB136 recipient strain; this marker lies on the side of phe distal to nic (Hoch & Matthews, 1973). Information concerning the disposition of the phe and nic genes on the purified B. subtilis DNA fragment was obtained by cleaving pSC5 with a variety of restriction endonucleases and then using the digested DNA to transform apheAl nic-38 strain of B . subtilis. As a precedent for the use of this approach, Lataste et al. (1981) have shown, using restriction endonucleasecleaved DNA, that the residual transforming activity for mutations in the amiA locus of Streptococcus pneumoniae depends on the distance of the mutations from the restriction endonuclease cleavage site. The uncut plasmid has an equal transformation efficiency for both genetic markers (Table 3). When the B. subtilis DNA was released from the plasmid by double digestion with EcoRI and BamHI, the transformation efficiency for both markers was reduced about tenfold. Cleavage with EcoRI alone reduced the number of Phe+ transformants to the same extent, which suggests that the pheA gene is at the end of the fragment that has the EcoRI Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 Integration of heterologous DNA in B. subtilis 1501 Table 2. Transforming activity of pSC5 for markers in the pheA region of the B. subtilis chromosome Recipient strain psc5 MB5 1 pheA I2 nic-38 phe+ psc5 168 J H 648 MB5 1 pheAl spoOB136 phe+ 0 5.4 4.3 (0% Spo+)* pheAI2 nic-38 168 JH648 pheAl spoOB136 phe+ nic+ phe+ 0.9 2.3 2.5 (70% Spo+)* Genotype Selected marker 10-5 x NO. of transformants (pg DNA)-' Donor DNA nic+ * The spoOB136 marker is linked by transformation to pheAI; Spo+ transformants were recognized by their brown colony pigmentation. Table 3. Eflect of cleavage with EcoRI, BamHI and SalGI on transforming activity of pSC5 against a pheAI nic-38 recipient strain Strain SA103 trpC2pheAI nic-38 was used as recipient. lops x No. of transformants (pg DNA)-' with donor: Selected marker psc5 psc5 (EcoRI) pscs (BamHI) pscs (EcoRI BarnHI) + psc5 (SulGI) phe+ nic+ 3.3 3.3 0.2 1.9 2.2 1.2 0.3 0.5 0-03 0.3 1 site. The number of Nic+ transformants obtained after cleavage with BamHI was slightly lower than the number obtained after cleavage with EcoRI but in neither case was the reduction in number as great as that observed after double digestion with both enzymes. This suggests that the nic gene lies near the centre of the fragment. Cleavage with SalGI, which cuts both the pBR322 vector and the B. subtilis insert once (see Fig. l), drastically reduced the number of Phe+ transformants obtained, suggesting that this restriction site lies relatively close to the position of the pheA mutation. Digestion with SalGI reduced the number of Nic+ transformants about tenfold. These results have been used to assign tentative positions to the pheA and nic genes on the restriction map of pSC5 (Fig. 1). Isolation of CmR transformants of B. subtilis The two plasmids pSC5 and p1949 were digested with EcoRI and then with BamHI. This released the 3.2 MDal phe-nic fragment from pSC5 and a 3.4 MDal fragment from p1949 carrying the CmR gene. The digested plasmids were mixed, ligated and used to transform a pheAI strain of B. subtilis. The numbers of Phe+ and CmRtransformants obtained are given in Table 4. As observed previously, cleavage of pSC5 with EcoRI and BamHI reduced the number of Phe+ transformants obtained by about two orders of magnitude. After ligation in the presence of restriction fragments derived from p1949, the number of Phe+ transformants obtained increased about tenfold. This presumably reflects the generation of larger DNA fragments which may afford a degree of protection against nucleolytic degradation during DNA uptake and recombination. As expected from the reports that p1949 is unable to replicate autonomously in B. subtilis (Haldenwang et al., 1980; Wilson et al., 1981), no CmRtransformants were obtained when p1949 DNA was used as donor (Table 4). However, after restriction and ligation with the fragments derived from pSC5, CmRtransformants were obtained, albeit at a relatively low yield. Ten transformants were picked at random; they were purified by streaking for single colonies on nutrient agar containing chloramphenicol and tested for their phenylalanine requirement. In view of the data in Table 2 which shows that pSC5 apparently harbours an incomplete copy of the pheA gene, it was expected that a mixture of phenylalanine auxotrophs and phenylalanine Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 1502 M. YOUNG Table 4. Isolation of CmR transformants of B. subtilis using as donor a ligated mixture of EcoRI plus BamHI-generated restriction fragments of pl949 and pSC.5 Strain JH648 trpC2pheAI spoOB136 was used as recipient. With E. coli strain C600 r- m- as recipient, p1949 gave 2.8 x lo5 CmRTcR transformants (pg DNA)-'. No.of transformants (pg DNA)-' I Donor pscs p 1949 Restricted pSC5 + restricted p1949* Phe+ CmR 5.6 x 1 0 5 ND ND Before ligation After ligation 3.8 x 103 2.7 x 104 I 0 0 3.8 x lo2 ND, Not determined. * Both plasmids were restricted with EcoRI and BamHI. Table 5. Three factor transduction crosses to determine the position of cat relative to pheA12 and nic-38 in strain CL104 The donor was CL104.2 trpC2 cat and the recipient was MB26 leu-8 pheAI2. Selected marker Recombinant class phe+ Leu- CmS Leu+ CmS Leu- CmR Leu+ CmR leu+ cat Phe- CmS Phe+ CmS Phe- CmR Phe+ CmR Leu+ Phe+ Leu- Phe+ Leu+ PheLeu- Phe- Frequency :li 20 Suggested order leu-phe-cat 68 23) leu-phe-cat 551 76 '1 leu-phe-cai 29 prototrophs would be obtained (see Fig. 4). However, all 10 strains (designated CL101-CL110) proved to be phenylalanine prototrophs. In accordance with this unexpected finding, when selection was made for Phe+ CmRdouble transformants using as donor the ligated mixture of double digested plasmids (see Table 4), the number of transformants obtained was very similar (data not shown) to that observed when the selection was for CmR transformants. Chromosomal location of the CmR determinant in strains CLlO1-CLllO DNA prepared from strains CL101-CL110 was used to transform E . coli strain C600 r- m-. None of the 10 DNA preparations yielded any CmRtransformants, whereas p1949 gave 2.8 x lo5 transformants (pg DNA)-'. This suggests that strains CL101-CL110 do not carry the CmR gene on an autonomous plasmid. Two further experiments were undertaken which establish the chromosomal location of the CmR determinant. In the first, the DNA preparations mentioned above were used at a nonsaturating concentration (10 ng DNA ml-I) to transform a pheAl strain of B. subtilis. In a representative experiment with strain CL104 as donor, Phe+ transformants arose at a frequency of 2.0 x lo6 (pg DNA)-' and only 4% of them were simultaneously transformed for the CmR marker. CmRtransformants arose at a reduced frequency of 2.2 x lo5 (pg DNA)-', but 97 % of them were simultaneously transformed to Phe+. This marked asymmetry has been observed previously (Wilson et al., 1981); presumably it reflects the presence in the donor strain of p1949 sequences for which no homology exists in the recipient strain. This asymmetry was not observed in the second experiment in which 3BS1 transduction was used to effect the transfer of Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 Integration of heterologous D N A in B. subtilis 1503 larger DNA fragments than are transferred during transformation. By this means we were able to demonstrate the linkage of the CmRdeterminant to the leu-8 marker, which was not involved in construction of the CL strains. The data from a representative experiment, using a Spo+ derivative of strain CL104 as donor, are given in Table 5, and they establish that the CmR marker is inserted into the B . subtilis chromosome on the side of pheAI2 distal to leu-8. Tests for linkage by transduction to a range of other genetic markers scattered about the B. subtilis chromosome indicated that insertion had occurred only in the pheA region. Similar results were obtained with several other CL strains (data not given). Strains CLIOI-I I0 produce Cms segregants As a working hypothesis it was postulated that the CmRdeterminant had been inserted into the bacterial chromosome by a Campbell mechanism. This would result in the generation of a duplication of thephe-nic fragment that was used to direct the insertion and hence it would be expected that the CmR strains would segregate CmS progeny at a measurable frequency. Moreover, since the incoming copy of the pheA gene was wild type, and the resident copy was mutant, it was expected that the CmS segregants would fall into two classes: those that were Phe+ and those that were Phe-. These predictions were tested by growing six of the 10 CL strains overnight in nutrient broth, and then plating them out on nutrient agar to obtain single colonies. Between 140 and 320 colonies derived from each strain were picked and replica plated on to (1) nutrient agar with and without chloramphenicol and (2) tryptophan minimal agar with and without phenylalanine. As expected, all the strains tested segregated CmS progeny at frequencies varying from 1.6% to 14.5%. However, none of the 150 CmS segregants that were tested, nor indeed any of the CmRparental types, proved to be phenylalanine auxotrophs. This therefore raised the question as to whether the pheAI mutation in the parental strain (JH648) was still present in the CL strains. Strain CLI04 does not appear to contain the pheAI mutation DNA from strain CL104 was used at a saturating concentration (0.1 pg ml-I) to transform strain MB21 of B. subtilis. Transformants to Met+, Leu+ and CmRwere selected separately using media that contained tryptophan and phenylalanine. This permitted the detection of double transformants that also carried the trpC2 or pheAl mutations presumed to be present in the donor strain. Transformants to Leu+ (160), Met+ (160) and CmR (320) were screened for the inheritance of the unselected donor markers. A total of 10 transformants were Trp-, but no Phecolonies were detected. This was particularly surprising in the case where selection had been made for CmR transformants since it had been established previously that when selection is made for this marker, approximately 97 % of the recombinants are simultaneously transformed for the donor phe marker. The implication of this experiment is that the donor strain no longer carries the pheAl mutation. The fact that the CL strains readily segregate CmS progeny suggests that they are merodiploid for the pheA-nic fragment and further evidence to confirm this was obtained using two approaches, namely genetic analysis and Southern hybridization. Genetic analysis of strain CLI04 Trp+ transformants of strain CL104 were obtained using as donor a nic-38 pheA I2 rif-2 strain of B . subtilis. The Ephrati-Elizur (1968) technique was used in this experiment since it generally results in higher frequencies of congression of unlinked donor markers than is observed with purified donor DNA (Nester et al., 1963) (see Table 6). Chloramphenicol was added to the selective medium to counterselect against growth of the donor strain. Further analysis of the Trp+ transformants showed that simultaneous transformation to Phe-, R i p and Spo+ had occurred at frequencies of 6-5%, 10.6% and 12.5%, respectively. In contrast, however, none of the Trp+ transformants had inherited the Nic- phenotype of the donor strain. In the control experiment, using as recipient a strain of B. subtilis that did not harbour the integrated CmR determinant, Trp+ Nic- double transformants arose at a frequency comparable to that observed for the Trp+ Phe-, Trp+ R i p and Trp+ Spo+ classes (Table 6). In the control cross it was not possible to counterselect against growth of the donor strain since the recipient was CmS. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 1504 M . YOUNG Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 Fig. 2. Southern hybridization experiment to identify sequences homologous to p1949 and the purified phe-nic fragment in restriction digests of D N A from strains CL104 and 168. (a) Restricted D N A samples before transfer to the nitrocellulose membrane. The restricted D N A samples were as follows: lane 1, 168 restricted with EcoRI; lane 2, p1949 with EcoRI plus BarnHI; (digestion incomplete) lanes 3-6, CL104 with EcoRI plus BamHI, EcoRI, BamHI, and SufGI, respcctively; lane 7 , 1 with EcoRI; lanes 8-10, CL104 with EcoRI, BamHI, and SafGI, respectively; lanes 11-13, 168 with SalGI, EcoRI, and BamHI, respectively; lane 14, I68 with EcoRI plus BamHI; lane 15, CL104 with EcoRI plus BamHI. (b) Autoradiogram of hybridized membrane filter. Lanes 1-6 were hybridized against p1949 DNA, lane 7 against 1DNA and lanes 8-15 against the purified phe-nzc fragment. 1506 M . YOUNG Table 6. Congression of rf-2, spoOB136, pheAl2 and nic-38 markers into strains CL104 and SAlOl Cross Recipient Donor CL104 trpC2 spoOB136 cat MB51 nic-38 pheAI2 rij12* Selected marker$ Recombinant class Frequency (%) Trp+ R i p 10.6 Trp+ Spo+ 12.5 Trp+ Phe6.5 Trp+ Nic0 2 SAlOl trpC2 spoOBl36 MB51 nic-38 pheAl2 trp+ Trp+ R i p 1.7 rij12t Trp+ Spo+ 1 .o Trp+ Phe1 .o Trp+ Nit1 .o * The spontaneous transformation technique of Ephrati-Eliiur (1968) was used. 7 A saturating concentration of purified DNA was used as donor. A total of 480 Trp+ transformants from each cross were tested for inheritance of the non-selected spo, rij; phe and nic markers. 1 trp+ Table 7. Eficiency of' transformation of pheA12, pheA1 and nic-38 recipients with strain CL104 as donor Cross Recipient 1 MB51 nic-38 pheAl2 2 MB51 nic-38pheAl2 3 SA 103 nic-38 pheA 1 4 SA 103 nic-38 pheA1 Donor CL104 trpC2 spoOB136 cat 168 trpC2 CL104 trpC2 spoOB136 cat 168 trpC2 Selected marker nic+ phe+ nic+ phe+ nic+ phe+ nic+ phe+ t No. of recombinants (pg DNA)-' 1.6 x 10" 2.8 x lo3 1.2 x 105 4-7 x 103 7.0 x lo5 5.1 x lo4 2.2 x lo5 2.0 x lo1 Ratio of Nic+/Phe+ transformants 5.6 2.5 1.4 1.1 Therefore, purified DNA was used as donor, and this accounts for the fact that transformation of the unselected donor markers occurred at a lower frequency than in the first cross in Table 6. These results, showing that CL104 is not readily transformed to Nic-, would be expected if the strain harbours two (or more) copies of the wild-type nic gene. The putative merodiploid strain CL104 was then used as donor in further crosses to measure the transformation efficiency against recipients carrying the pheA12, pheAl and nic-38 markers. It should be noted that the putative merodiploid region in strain CL104 does not include the sequences complementary to the pheAl2 mutation (see Table 2). A non-saturating concentration of DNA from strain CL104 was used to transform apheAl2 nic-38 recipient strain; DNA from strain 168 was used as control (Table 7, crosses 1 and 2). Transformants to Phe+ and to Nic+ were selected separately, and the frequencies obtained were used to calculate the ratio of Nic+/Phe+transformants with each donor. When the donor was strain 168 (cross 2), the ratio was approximately 2.5, which presumably reflects a difference in the transformation efficiencies for the nic-38 and pheAl2 markers. When the donor was strain CL104, the ratio was increased about twofold to 5.6. This suggests that sequences complementary to the nic-38 mutation are present in strain CL104 in approximately twice as many copies as are to be found in normal 168 strains. This supports the contention that the strain is merodiploid for the nic region. Evidence that the merodiploid region also includes the sequences complementary to the pheAl mutation was obtained in crosses 3 and 4 in Table 7. In these crosses the same donor strains were used to transform a pheAl nic-38 recipient. As previously, Phe+ and Nic+ transformants were selected separately, and the frequencies observed were used to calculate the ratio of Nic+/Phe+ transformants. This ratio was close to unity with strain 168 as donor and was only slightly increased with strain CL104 as donor. However, it is noteworthy that the transformation efficiencies for both markers are 2.5-3-fold greater with strain CL104 as donor than with strain Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 Integration of heterologous DNA in B. subtilis B 5 E S m B m E B E WA 1507 S 1 Fig. 3. Restriction map of thephe-nic region of the chromosome in strain CL104. The figure is based on the data given in Fig. 2. The inserted p1949 sequences are represented by a single line. Regions corresponding to the pheA and nic genes are shown diagrammatically by solid and hatched areas, respectively. Although there is evidence that the 6.6 MDal inserted D N A fragment may be tandemly duplicated in strain CL104 (see text), this is not shown in the figure. The numbers represent MDal. B, B a m H I ; E, EcoRI; and S, SalGI. 168 as donor. These results indicate that sequences complementary to the pheAI and nic-38 mutations are present in greater abundance in strain CL104 than in strain 168, but that their abundance relative to each other remains similar. Restriction map of the phe-nic region of strain CL104 DNA from strains CL104 and 168 was cleaved with a fivefold excess of EcoRI, BamHI and SalGI. The restriction fragments were electrophoretically separated in agarose according to their molecular weight and transferred to a nitrocellulose membrane. The membrane was then cut into strips and the appropriate samples hybridized against (1) nick-translated p1949, or (2) nicktranslated 3.2 MDal phe-nic fragment of pSC5 (see Fig. 1). A molecular weight standard of A DNA cut with EcoRI was also included. The hybridized and washed membrane filter strips were then reassembled and autoradiographed and the results are given in Fig. 2. When p1949 was used as probe against strain CL104 DNA cleaved with EcoRI plus BamHI (lane 3) and against p1949 DNA cut with the same enzymes (lane 2), hybridization occurred with a 3.4 MDal fragment corresponding to the EcoRI-BamHI fragment of p1949 that carries the CmR determinant (see Fig. 1). When CL104 DNA was restricted with either EcoRI, BamHI or SalGI the probe hybridized to a 6.6 MDal fragment (lanes 4-6). This corresponds to the sum of the 3.4 MDal EcoRI-BamHI fragment of p1949 and the 3.2 MDal phe-nic fragment of pSC5 (see Fig. 1). There was no hybridization of the probe to EcoRI-cleaved DNA from strain 168 (lane 1). In lanes 8-15 the 3.2 MDal phe-nic fragment of pSC5 was used as hybridization probe. Hybridization to CL104 DNA restricted with either EcoRI, BamHI or SalGI showed the same 6.6 MDal fragment as was detected with the p1949 probe, but in addition each enzyme also produced a second fragment of higher molecular weight to which much weaker hybridization occurred (lanes 8-10). When hybridized against strain 168 DNA that had been restricted with the same enzymes, only the higher molecular weight bands appeared (lanes 11-13). Finally, when either strain 168 or strain CL104 DNA was cut with both EcoRI and BamHI (lanes 14 and 15), hybridization occurred with a 3.2 MDal fragment corresponding to the 3.2 MDal phe-nic fragment of pSC5 (see Fig. 1). These data were used to construct the restriction map of the phe-nic region of the CL104 chromosome shown in Fig. 3. DISCUSSION The results obtained in this investigation confirm previous reports (Duncan et al., 1978; Haldenwang et al., 1980; Wilson et al., 1981 ; Niaudet et al., 1982)that heterologous DNA can be inserted into the B . subtilis chromosome provided that the incoming D N A has some sequence homology with the bacterial chromosome. Moreover, they demonstrate that insertion occurs into that region of the chromosome from which the exogenote was derived. In none of the 10 independent CmR transformants studied here was there any evidence that there were sites of insertion into the bacterial chromosome other than the site in thepheA region from which the Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 1508 M . YOUNG Region I Region B pheA I Fig. 4. Diagrammatic representation of the integration of the p1949 CmR determinant into the pheA region of the B . subtilis chromosome by a Campbell mechanism. The heterologous p1949 sequences (3.4 MDal) are shown as a single line and the phe-nic fragment (3.2 MDal) as a double line. The pheA and nic genes are represented by solid and hatched areas, respectively. The recipient strain (JH648) carries thepheAZ mutation, which lies close to the EcoRI site within thepheA gene (Table 2). Since the 3.2 MDal fragment used to direct insertion carries an incomplete copy of the pheA gene, integration by a Campbell mechanism could generate either Phe+ or Phe- recombinants, depending on whether the recombination event occurs to the left (region A) or right (region B) of thepheAl mutation, respectively. B, BarnHI and E, EcoRI. exogenote was derived. Also, none of the transformants harboured an autonomous plasmid. Many of the properties of the CmR transformants are consistent with the occurrence of integration by a Campbell recombination mechanism. Three independent experiments indicated that the strains are merodiploid for the sequences used to promote insertion of the CmR determinant. The CmR transformants readily gave rise to CmS segregants. When using pheAl and nic-38 recipients, DNA from the CmRtransformants gave about twice as many Phe+ and Nic+ transformants as did DNA from the wild type strain (Table 7). Southern hybridization experiments indicated that the heterologous plasmid DNA was bounded on both sides by copies of the phe-nic fragment (Fig. 2). Since the exogenotic phe-nic fragment of B. subtilis DNA apparently carries an incomplete copy of the wild type pheA gene (Table 2) it was anticipated that integration by a Campbell mechanism would generate a mixture of Phe+ and Phe- transformants, depending on the precise location of the recombination event (see Fig. 4). Contrary to expectation, all of the CmR transformants tested were Phe+. Assuming for the present that integration was by a Campbell mechanism, then in order to account for this finding it must be proposed that the likelihood of a crossover occurring within the pheA gene, between the EcoRI end of the cloned fragment and the site of the pheA1 mutation, is much greater than the likelihood of a crossover occurring between the BamHI end of the cloned fragment and the site of the pheAl mutation (see Fig. 4). This proposal seems unlikely in view of the fact that the pheAl mutation seems to lie relatively close to the EcoRI end of the cloned fragment (Table 2). However, it could be accounted for if there was a recombination hot spot (chi sequence?) (Stahl, 1979; Smith et al., 1981) in this region. There is, however, another serious difficulty with the proposal that integration occurred by a Campbell type of mechanism: namely, that the CmRtransformants do not appear to harbour the pheAl mutation. Experimentally it was found that the pheAI mutation, supposedly present, could not be rescued from the strains by transformation. Also, when a large number of CmS segregants were obtained from the CmR transformants, all of them were Phe+. Finally, it was demonstrated that one of the CmRtransformants (CL104)carries at least two wild type copies of the sequences complementary to the pheA2 mutation (Table 7). Bresler et al. (1968) and Spatz & Trautner (1970) have shown that during the processes of transformation and transfection in B . subtilis, allelic conversion can occur. DNA heteroduplexes, containing wild type and mutant copies of a gene may be converted to homoduplexes before replication leads to the partitioning of the two allelic forms of the gene in the resultant Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 Integration of heterologous DNA in B. suhtilis 1509 Fig. 5. Model depicting integration of a linear ligation product carrying the p1949 CmR determinant into thepheA region of the B. subtilis chromosome. A copy of thephe-nic fragment is ligated to each end of the heterologous p1949 D N A (single line). Steric hindrance favours the occurrence of recombination near the extremities of the phe-nzc fragments. As a result the pheA J mutation in the recipient strain (JH648) is lost on integration. B, BamHI and E, EcoRI. daughter chromosomes. Although in the present investigation we are not dealing with DNA heteroduplexes, it might still be argued that a form of allelic conversion has occurred, since it has been shown that alterations introduced into the recipient-homologous DNA of hybrid plasmids can be corrected in transformation (Iglesias et al., 1981). Very recently Chak et al. (1982), have extended this finding by their demonstration of the in vivo transfer of mutations between the B . subtilis chromosome and a plasmid harbouring homologous DNA. If gene conversion is responsible for the results reported here, then it must be explained why the pheA2 endogenote has apparently been converted by the phe+ exogenote in all 10 transformants that were examined. The selective medium used to obtain the CmRtransformants was supplemented with phenylalanine and, as far as I am aware, pheAl strains of B. subtilis grow and sporulate normally under the conditions employed. Hence, no selective pressure was applied to force conversion of the mutantpheAI allele to the wild-typephe+ allele. It is noteworthy, in this context, that Spatz & Trautner (1970) did observe asymmetric gene conversion in their experiments on the transformation of B. subtilis with heteroduplex SPPl DNA. However, wild type and mutant alleles of a variety of different genes have been shown to coexist in merodiploid strains of B . subtilis (Trowsdale & Anagnostopoulos, 1975, 1976; Schneider & Anagnostopoulos, 1981 : Fink & Zahler, 1982). On balance therefore, it is improbable that the systematic loss of the pheA2 mutation observed here is the result of allelic conversion. An alternative trivial explanation for the failure to detect the pheA2 mutation in the merodiploid strains must also be considered. It seems very likely that multiple copies of the exogenote are present in the strains (this is discussed later). If the exogenote is present in very large excess, we may have failed to detect the single copy of the endogenote for the trivial reason that insufficient numbers of CmS segregants and transformants were analysed (see Results). This explanation might also account for the initial finding that all of the CmR transformants obtained with the ligated DNA were Phe+. In conclusion, the possibility that the exogenote integrated by a Campbell mechanism cannot be ruled out on the basis of the evidence currently available. An alternative mechanism for integration of the CmR determinant can, however, be envisaged. A reciprocal recombination event might have occurred between the bacterial chromosome and a linear ligation product comprising the 3.4 MDal plasmid fragment carrying the CmRdeterminant with a copy of the 3.2 MDalphe-nic fragment of B. subtilis DNA ligated to each extremity. Integration could then proceed by the formation of a bridge structure, as shown diagrammatically in Fig. 5. Steric hindrance might favour the occurrence of recombination in Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 15 10 M. YOUNG the region between the pheAl mutation and the EcoRI end of the fragment, thus accounting for the fact that the resultant merodiploid strains have apparently lost the pheA1 mutation. Evidently, i t dissimilar chromosomal fragments were ligated to the ends of the heterologous DNA, insertion by this mechanism would result in deletion of the chromosomal sequences that lie between them (Niaudet et al., 1982). A relatively simple method for the generation of deletion mutations, for which positive selection could be made, could be developed by the use of p1949 as an insertion vector, and refinements enabling the production of deletions of varying extent from the ends of fragments already cloned in the insertion vector can also be envisaged. A recombination mechanism similar to that described here has been proposed for the transformation of a large deletion mutation in B. subtilis (Adams, 1972), and also for the transformation of a heterologous intergenote comprising the ‘Bacillusglobigii’ aromatic region in B. subtilis (Harris-Warwick & Lederberg, 1978). It would be of interest to determine whether the integration of a plasmid comprising the 3.4 MDal CmRfragment of p1949 and the 3-2 MDalphe-nic fragment does in fact proceed by a Campbell mechanism. Repeated attempts to construct such a plasmid by ligating the two purified fragments and transforming either a recBC derivative of E. coli strain C600 or a recA strain (HB101) have failed. As an alternative approach, EcoRI or BamHI digests of CL104 DNA have been ligated and used to transform the strains of E. coli mentioned above, on several different occasions. However, it has not proved possible to obtain the desired product. It seems that a plasmid comprising the 3.4 MDal p1949 sequences plus the 3.2 MDalphe-nic fragment is not stably maintained in E. coli. This is made all the more difficult to understand since the 3.2 MDal phe-nic fragment is stably maintained in E. coli when it is cloned into pBR322 (J. Szulmajster & C. Bonamy, personal communication). In view of these unexpected difficulties, it has not yet proved possible to demonstrate that p1949 sequences can integrate into the B . subtilis chromosome by a Campbell mechanism when they are carried on a covalently closed plasmid. Finally, it is of interest that in the Southern hybridization experiment (Fig. 2, lanes 8-10) using the phe-nic fragment as probe, hybridization occurred to two restriction fragments of CL104 DNA, but with very disparate intensities. The 6-6 MDal fragment comprising the vector plusphe-nic segment showed much more intense hybridization than did the larger fragment, also seen in digests of DNA from the wild type strain (lanes 11-13). This presumably reflects differences in the numbers of copies of the two fragments present in the digested DNA, the implication being that the vector plusphe-nic fragment is present in multiple copies (probably at least 10) in strain CL104. (The transformation analysis discussed above suggested that only about two copies were present, and the reason for the discrepancy is not known.) This phenomenon has previously been noted by Wilson et al. (1 98 1) who employed the same plasmid as used here, and also by Tanaka (1979) who reported that the leu region of the B . subtilis chromosome undergoes spontaneous recE-independent duplications when it is carried on an autonomously replicating plasmid. Independent evidence confirming that multiple copies of the inserted sequences are present was obtained by digesting DNA from the 10 CmR strains (CL101-110) with EcoRI and BamHI and examining the restriction digests by agarose gel electrophoresis. In each case a more or less prominent band was discernible, with an apparent molecular weight of about 6.7 MDal against the background of restriction fragments generated from the chromosomal DNA (M. Young, unpublished results). Moreover, this band was converted into two bands of apparent molecular weights of about 3-5 and 3.2 MDal after double digestion with EcoRI and BamHI. In some of the CmR transformants the band was more prominent than in others, but no correlation was observed between band intensity and the maximum chloramphenicol tolerance of the strains. It is pertinent to ask where these multiple copies of the inserted sequences reside. Since we were unable to detect free plasmids in the strains, and there was no evidence for insertion occurring anywhere else than in thepheA region, it may be concluded that the inserted sequences are probably tandemly duplicated in the pheA region. Evidence to confirm this hypothesis is currently being sought. It is possible that growth in the presence of chloramphenicol provides sufficient selective pressure to force duplication of the inserted sequences in situ. Alternatively, they may have arisen by multiple integration events in the initial transformation. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 Integration of heterologous D N A in B. subtilis 1511 In any event, the finding that multiple copies of the cloned B. subtilis DNA are present in the strains makes it much less probable that this particular insertion vector will prove to be particularly useful for the analysis of dominance and complementation relationships of mutations in B . subtilis. Whether or not other insertion vectors will show the same behaviour remains to be seen. Note added in proof: Ferrari et al. (1982) have recently published a restriction map of the pheA region of the B. subtilis chromosome which confirms the results reported here. I thank Dr J. Szulmajster, in whose laboratory this work was started, for his hospitality during a three month period when I was the recipient of an EMBO Short Term Fellowship. I also thank D r M. G. Sargent for his help in the preparation of nick-translated DNA, Mr G. Price for his technical assistance, and Miss S. W. Evans for typing the manuscript. The financial support of the SERC (GR/B75082) is gratefully acknowledged. REFERENCES ADAMS,A. (1972). Transformation and transduction of FERRARI, F. A., LANG,D., FERRARI,E. & HOCH,J. A. a large deletion mutation in Bacillus subtilis. Molecu(1 982). Molecular cloning of the s p d B sporulation lar and General Genetics 118, 31 1-322. locus in bacteriophage lambda. Journal of BacterioANAGNOSTOPOULOS, C. & SPIZIZEN, J. (1961). Requirelogy 152, 809-814. ments for transformation in Bacillus subtilis. Journal FINK, P. S. & ZAHLER, S. A. (1982). Specialised of Bacteriology 81, 741-746. transduction of the ihD-thyB-ilcA region mediated ANDERSON, L. M., RULEY,H. E. & B o n , K . F. (1982). by Bacillus subtilis bacteriophage SPO. Journal of Isolation of an autonomously replicating DNA Bacteriology 150, 1274-1 279. fragment from the region of defective bacteriophage G. & ALBERGALIZZI,A., SCOFFONE,F., MILANESI, PBSX of Bacillus subtilis. Journal of Bacteriology 150, TINI,A. M. (1981). Integration and excision of a 1280-1286. plasmid in Bacillus subtilis. Molecular and General BRESLER, S. E., KRENEVA, R. A. & KUSHER,V. V. Genetics 182, 99-105. GRANT,W . D. (1974). Sporulation in Bacillus subtilis (1 968). Correction of molecular heterozygotes in the course of transformation. Molecular and General 168. Control of synthesis of alkaline phosphatase. Genetics 102, 257-268. Journal of General Microbiology 82, 363-369. CAMPBELL,A. M. (1962). Episomes. Aduances in HALDENWANG, W . G., BANNER, C. D. B., OLLINGTON, Genetics 11, 101-145. J . F., LOSICK,R., HOCH,J. A., O’CONNOR,M. B. & CHAK, K. F., DE LENCASTRE, H., LIU,H-M. & PIGGOT, SONENSHEIN, A. L. (1980). Mapping of a cloned gene P. J. (1982). Facile in cico transfer of mutations under sporulation control by insertion of a drug between the Bacillus subtilis chromosome and a resistance marker into the Bacillus subtilis chromoplasmid harbouring homologous DNA. Journal of some. Journal of Bacteriology 142, 90-98. General Microbiology 128, 28 13-28 16. J. (1978). HARRIS-WARWICK, R. M. & LEDERBERG, CHAN,C. S. M. & TYE,B. K . (1980). Autonomously Interspecies transformation in Bacillus: mechanism replicating sequences in Saccharomyces cereuisiae. of heterologous intergenote transformation. Journal Proceedings of the National Academy of Sciences ojthe of Bacteriology 133, 1246-1253. United States of America 77, 6329-6333. HARRIS-WARWICK, R. M., ELKANA, Y., EHRLICH,S. D. D. R. (1969). Supercoiled CLEWELL, D. B. & HELINSKI, & LEDERBERG, J. (1975). Electrophoretic separation circular DNA-protein complex in Escherichia coli: of Bacillus subtilis genes. Proceedings of the National purification and induced conversion to an open Academy of Sciences of the United States of America circular DNA form. Proceedings of the National 72, 2207-22 1 1. Academy of Sciences of the United States of America HOCH,J. A. & MAITHEWS,J. L. (1973). Chromosomal 62, 1159-1 166. location of pleiotropic negative sporulation mutations in Bacillus subtilis. Genetics 73, 21 5-228. H. S. DONELLAN, J. E. JR, NAGS,E. H. & LEVINSON, HORINOUCHI, (1964). Chemically defined, synthetic media for S. & WEISBLUM, B. (1982). Nucleotide sequence and functional map of pC194, a plasmid sporulation and for germination and growth of that specifies inducible chloramphenicol resistance. Bacillus subtilis. Journal of Bacteriology 87, 332-336. Journal of Bacteriology 150, 8 15-825. DUGAICZYK, A,, BOYER,H. W. & GOLDMAN, H. M . HRANUELI,D., PIGGOT,P. J. & MANDELSTAM, (1975). Ligation of EcoRI endonuclease-generated J. DNA fragments into linear and circular structures. (1974). Statistical estimate of the total number of Journal of Molecular Biology 96, 171-1 84. operons specific for Bacillus subtilis sporulation. DUNCAN, C. H., WILSON, G. A. &YOUNG,F. E. (1978). Journal of Bacteriology 119, 684 690. Mechanism of integrating foreign DNA during IGLESIAS, A., BENSI,G., CANOSI, U . & TRAUTNER, T . A. transformation of Bacillus subtilis. Proceedings of the (198 1). Plasmid transformation in Bacillus subtilis. National Academy of Sciences of the United Stares of Alterations introduced into the recipient-homoloAmerica 75, 3664-3668. gous DNA of hybrid plasmids can be corrected in EPHRATI-ELIZUR, E. (1968). Spontaneous transtransformation. Molecular and General Genetics 184, formation in Bacillus subtilis. Genetical Research 11, 405-409. 83-96. IORD~NESCU, S. (1975). Recombinant plasmid ob- Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35 15 12 M. YOUNG tained from two different, compatible staphylococcal plasmids. Journal of Bacteriology 124, 597-601. JEFFRIES, A . J . & FLAVELL,R. A . (1977). A physical map of the D N A regions flanking the rabbit 8-globin gene. Cell 12, 429439. KARAMATA, D. & GROSS,J . D . (1970). Isolation and genetic analysis of temperature-sensitive mutants of B. suhrilis defective in D N A synthesis. Molecular and General Genetics 108, 277-287. KAWAMURA, F . , SAITO.M. & IKEDA, Y . (1979). A method for construction of specialised transducing phage pl 1 of Bacillus suhtilis. Gene 5, 87-91. LATASTE, H., CLAVERYS, J-P. & SICARD,A . M. (198 1). Relation between the transforming activity of a marker and its proximity to the end of the D N A particle. Molecular and General Genetics 183, 199201. MARMUR,J . (1961). A procedure for the isolation of deoxyribonucleic acid from micro-organisms. Journal of Molecular Biology 3, 208-218. MANDEL,M . & HIGA,A . (1970). Calcium-dependent bacteriophage D N A infection. Journal o j Molecular Biology 53, 159-162. MEJEAN,V., CLAVERYS, J-P., VASSEGHI, H. & SICARD, A-M. (1981). Rapid cloning of specific D N A fragments of Streptococcus pneumoniae by vector integration into the chromosome followed by endonucleolytic excision. Gene 15. 289-293. MESSER.W., BERGMANS.H . E. N., MEIJER,M., WOMACK, J . E., HANSEN,F. G . & VON MEYENBURG, K . (1978). Minichromosomes: plasmids which carry the E. coli replication origin. Molecular and General Genetics 162, 269-275. NESTER,E. W., SCHAFER,M. & LEDERBERG, J . (1963). Gene linkage in D N A transfer: a cluster of genes concerned with aromatic biosynthesis in Bacillus subtilis. Genetics 48, 529-55 1. NIAUDET, B. & EHRLICH, S. D. (1979). In uitro genetic labelling of Bacillus suhtilis cryptic plasmid pHV400. Plasmid 2, 48-58. NIAUDET,B., GOZE, A. & EHRLICH,S. D. (1982). Insertional mutagenesis in Bacillus subtilis : mechanism and use in gene cloning, Gene 19, 277-284. OESCHGER, M. P., OESCHGER,N . S., WIPRUD,G . T. & WOODS, S. L. (1980). High efficiency temperaturesensitive amber suppressor strains of Escherichia coli K 12 : isolation of strains with suppressor-enhancing mutations. Molecular and General Genetics 177, 545552. PIGGOT, P. J. (1973). Mapping of asporogenous mutations of Bacillus subtilis; a minimum estimate of the number of sporulation opcrons. Journal of Bacteriology 114, 1241-1253. RAPOPORT, G., KLIER,A., BILLAULT, A., FARGETTE, F. & DEWNDER, R. (1979). Construction of a colony bank of E. coli containing hybrid plasmids represen- tative of the Bacillus subtilis 168 genome. Moleculur and Generul Genetics 176, 239-245. RUBIN,E. M., WILSON,C. A . & YOUNG,F . E. (1980). Expression of thymidylate synthetase activity in Bacillus suhtilis upon integration of a cloned gene from Escherichia coli. Gene 10, 227-235. SAGER,R., ANISOWICZ,A . & HOWELL,N . (1981). Genomic rearrangements in a mouse cell line containing integrated SV40 D N A . Cell 23, 41-50. SCHNEIDER, A.-M. & ANAGNOSTOPOULOS, C. (1981). Linkage maps and properties of a Bacillus suhtilis strain carrying a non-tandem duplication of the purB-trc. region of the chromosome. Journal of’ General Microbiology 125, 241 -256. SMITH, G . R . , KUNES,S. M., SCHULTZ, D. W., TAYLOR, A . & TRIMAN, K . L. (1981). Structure of chi hotspots of generalised recombination. Cell 24, 429-436. SOUTHERN,E. M. (1975). Detection of specific sequences among D N A fragments separated by gel electrophoresis. Journal of Molecular Biology 98, 503-5 17. SPATZ,H . C. & TRAUTNER, T . A . (1970). One way to do experiments on gene conversion : transfection with heteroduplex SPPl D N A . Molecular and General Generics 109, 84-106. STAHL, F . W. (1979). Special sites in generalised recombination. Annual Review of Genetics 13, 7-24. TAKAHASHI, 1. (1963). Transducing phages for Bacillus suhrilis. Journal oJ’Genera1Microbiology 31. 2 1 1-2 1 7. TANAKA, T. (1979). recE4independent recombination between homologous deoxyribonucleic acid segments of Bacillus suhtilis plasmids. Journal uf Bacteriology 139, 775-782. TROWSDALE,J . & ANAGNOSTOPOULOS, C. ( 1975). Evidence for the translocation of a chromosome segment in Bacillus subtilis strains carrying the trpE26 mutation. Journal of’ Bactcriologj, 122, 886898. TROWSDALE, J . & ANAGNOSTOPOULOS, C. (1976). Differences in the genetic structure of Bacillus suhtilis strains carrying the trpE26 mutation and strain 168. Journal qf Bacteriology 126, 609-61 8. WAHL, G . M., STERN,M. & STARK,G. R . (1979). Efficient transfer of large D N A fragments from agarose gels to diazobenzyloxymethyl-paper and rapid hybridisation by using dextran sulfate. Proceedings of’ the National Academy yf Scicwces of’ the United States qf’ America 76, 3683 3687. WILSON,F. E., HOCH,J . A . & BOTT,K . (1981). Genetic mapping of a linked cluster of ribosomal ribonucleic acid genes in Bacillus suhtilis. Journul of’Bactt’rio1og.v 148, 624-628. YONEDA,Y . , GRAHAM,S. & YOUNG, F. E. (!979). Cloning of a foreign gene coding for Ir-amylase in Bacillus subrilis. Biochemical and Biophj..yicalResearch Coniniunicatioits 91, 1556- 1564. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sun, 18 Jun 2017 18:31:35