A new cloning system for Bacillus subtilis comprising elements of phage, plasmid and transposon vectors

Gene, 73 (1988) 215-226 Elsevier

215

GEN 02744

A new cloning system for Bacillus subtilis comprising elements of phage, plasmid and transposon vectors (Recombinant DNA; bacteriophage SP/I; prophage transformation;

Tn917; genomic library)

Harold Poth * and Philip Youngman Department of Microbiology. University of Pennsylvania School of Medicine, Philadelphia, PA 19104 (U.S.A.) Received 18 May 1988 Revised and accepted 12 August 1988 Received by publisher 30 August 1988

SUMMARY

A new cloning system for Bacillus subtilis was devised which makes use of a combination of Tn92 7-containing phage SP/? derivatives and Tn917-containing Escherichia co&B. subtilis shuttle plasmids. This system allows the initial cloning of genes in single copy, via ‘prophage transformation’, with a selection for complementation of mutational defects in B. subtilis hosts and permits subsequent transfer of the cloned material by homologous recombination to low-copy and high-copy vectors that replicate in both B. subtilis and E. coli. Because cloned sequences are adjacent to pB322-derived DNA in the recombinant phages, inserts can also be ‘rescued’ directly from the phage DNA after digestion with appropriate restriction enzymes, circularization of the fragments by ligation and transformation of an E. coli recipient. Two genomic libraries of B. subtilti chromosomal Sau3Agenerated partial-digest fragments in the size ranges of 5-8 kb and 8-10 kb were constructed and screened for the complementation of mutations aroZ906, cysA 14, dal-1, ghB 133, metC3, purA 16, purB33, thrA5, trpC2 and recE4. In all cases, specialized transducing phages carrying inserts that complemented the selected markers were recovered. Inserts complementing the dal-1 and trpC2 mutations could be transferred from recombinant phages to Tn917-containing plasmids by homologous recombination without in vitro subcloning. Another insert complementing the purB33 mutation was rescued directly into E. coli from a recombinant phage DNA.

Increased interest in B. subtilis as an experimental model system and as a host for certain industrial

applications has already led to the development of several useful plasmid- or phage-mediated molecular cloning systems (reviewed by Errington, 1987). Existing systems have significant limitations, how-

Correspondenceto: Dr. P. Youngman, Department of Microbiology, Rm. 209 Johnson Pavilion, University of Pennsylvania School of Medicine, Philadelphia, PA 19104 (U.S.A.) Tel. (215)898-2887. * Present address: Department of Microbiology, Gesellschaft Btr Biotechnologie Forschung mbH, Mascheroder Weg 1, D-3300 Braunschweig (F.R.G.) Tel. (0531)6181-O.

Abbreviations: Ap, ampicillin; bla, gene coding for p-lactamase; cat, gene coding for Cm acetyltransferase; Cm, chloramphenicol; A, deletion; EMS, ethyl methanesulfonate; kb, kilobase pairs; MC, mitomycin C; MLS, macrolide, lincosamide and streptogramin B antibiotics; pBR, pBR322-derived sequences; R, resistance; ‘, sensitivity; Tc, tetracycline; TE, see MATERIALS AND METHODS, section c; rer, gene conferring Tc resistance; TSS, see MATERIALS AND METHODS, section b; wt, wild type; : :, novel joint.

INTRODUCTION

0378-I 119/88/$03.50

0 1988 Elsevier

Science Publishers

B.V. (Biomedical

Division)

216

ever. The cloning of chromosomal genes using multicopy plasmid vectors can be deleterious to the host or may cause structural or segregational instability of the recombinant plasmids (reviewed by Dubnau, 1983). The use of temperate bacteriophages like SP/?, pl 1 or $105 avoids these problems but introduces still others. The large size of SP/? or pll makes it possible to clone relatively large inserts, but complicates further analysis of the cloned material. The recently improved $105 vectors developed by Errington and coworkers (Errington and Jones, 1987; Jones and Errington, 1987) have been used successfully to clone many genes, but can accommodate only inserts less than 4 kb in size. Subcloning of inserts from recombinant $105 phages is less difficult than with SPP or ~11, but still can present technical obstacles. In the present work we have explored a new approach designed to minimize or eliminate problems inherent in the kinds of plasmid and phage vectors in current use for the cloning of B. subtilis chromosomal DNA in a B. subtilis host. This approach is based on the concept of ‘prophage transformation’ pioneered by Kawamura and colleagues (Kawamura et al., 1979) and makes use of a nondefective transposon-containing derivative of the large temperate phage SP/I (Zahler et al., 1982; Youngman et al., 1983). Its novelty consists in the fact that genomic libraries are obtained in a way that places cloned inserts in association with a cut gene selectable in B. subtih and adjacent to pBR322derived (pBR) sequences (Bolivar et al., 1977) that include the ColEl replication origin and blu (Blactamase) gene. This makes it possible to ‘rescue’ cloned inserts by cutting the recombinant phages or chromosomal DNA with certain restriction enzymes, circularizing the fragments by ligation and transforming directly into an E. coli recipient. Moreover, cloned inserts are flanked by Tn917 sequences in the recombinant phages. This makes it possible to transfer inserts from the phage DNA to a B. subtilis plasmid by homologous recombination. We describe here the construction of a first generation of vectors designed to test the effectiveness of this kind of Tn917-mediated prophage transformation and ‘recombinational subcloning’.

MATERIALS

AND METHODS

(a) Bacterial strains and plasmids Bacterial strains and plasmids used in this study are listed in Table I. (b) Culture media and genetic techniques LB broth, prepared as described previously (Youngman, 1987), was used as a complex medium for B. subtilis and E. coli supplemented with 50 pg/ml of D-alanine for Dal- B. subtilis strains. TSS medium, prepared as described previously (Youngman, 1987), was used as a minimal medium for B. subtilis, supplemented with required growth factors at a concentration of 25 pg/ml. Transformation of B. subtilis strains was carried out as described by Anagnostopoulos and Spizizen (1961) and transformation of E. coli strains was carried out as described by Cohen et al. (1973). Selections for antibiotic resistances were carried out as described previously (Youngman, 1987). Screening for complementation of the recE4 mutation in B. subtiliswas carried out by selection for resistance to mitomycin C (MC) (0.04 pg/ml) or ethyl methanesulfonate (EMS) (0.1%). Techniques for use of B. subtilis phage SP#l were essentially as described by Rosenthal et al. (1979). (c) In vitro manipulation of DNA Chromosomal DNA of B. subtilis was isolated as described previously (Guzman et al., 1988). The alkaline lysis procedure of Birnboim and Doly (1979) was used for the extraction of plasmid DNA from B. subtilisand E. coli. The SPfiphage DNA was isolated with a rapid small-scale procedure. Heat-induced lysates (35 ml) were centrifuged (Sorvall SS34 rotor, 10 000 rpm, 15 min) to remove unlysed cells and particulate debris and the phage particles were concentrated from the supematant by high-speed centrifugation (Beckman Ti70 rotor, 28 000 rpm, 1 h). The crude phage pellets were resuspended in 200 ~1 each of 10 mM NaCl, 50 mM Tris * HCl, pH 8.0, 10 mM MgCl,, lysed in the presence of 20 mM EDTA, 0.5% SDS and treated with proteinase K (50 pg/ml, 1 h, 65 “C). After at least two phenolchloroform extractions, DNA was concentrated by

217 TABLE I Strains and plasmids Strain or plasmid

Source or reference

Genotype or relevant characteristics

B. subtilis strains: attSPB+, Ret + , prototrophic PY79 sup-3 ade- leu- thr- metB5, SPbs cu1050 hpC2, attSP/I CU1065 metB5 cysA 14 purA 16 attSPb cu1441 metB5 aroI906 purB33 da/-l attSP/? cu1442 ilvD 15 tre-12 metC3 g!vB 133 attSPb cu1443 metB5 hisA 1 thrA5 attSPp CU1448 trpC2 (SPjc2dZ::Tn917) CU2530 metB5 t1pC2 M-1 recE4 attSPB YB1015 CU1050 (SP/?c2A2::Tn917::pTV56cut) PY466 PY79 (pTV5) PY468 PY79 (SP,%2A2::Tn917) PY480

Youngman et al. (1984) S.A. Zahler S.A. Zahler S.A. Zahler S.A. Zahler S.A. Zahler S.A. Zahler S.A. Zahler R.E. Yasbin This work This work This work

E. co/i strains: hsdS20 supE44 oral4 galK2 lacy1 proA rpsL20 xyl-5 mtl-1 recA 13 rglHBlOl

Boyer and R.-Dussoix (1969)

Plasmids: pBR322 pTV5 pTV13 pTV56 pcv1

Bolivar et al. (1977) Youngman et al. (1984) H.-P. Biemann, unpublished This work This work

ColEl replicon containing blu and tet genes (Fig. 1) Tn9I7-containing pE194 derivative (Fig. 4) pTV5 derivative in which a cut gene has been inserted into Tn.917 sequences (Fig. 1) pTV13-pBR322 chimera (Fig. 1) prophage transformation cloning vector (Fig. 2)

ethanol precipitation and finally resuspended in TE (10 mM Tris * HCl, pH 8.0, 1 mM EDTA).

RESULTS AND DISCUSSION

(a) Construction of shuttle vector pTV56

Cloning by ‘prophage transformation’ typically involves the ligation of random fragments of genomic DNA to random or specific fragments of phage DNA, followed by transformation of a host lysogenic for that phage (Kawamura et al., 1979; Saava and Mandelstam, 1984; Seki et al., 1986). The ligated phage sequences allow genomic fragments to become integrated into the resident prophage by homologous recombination. Our intention was to adopt essentially the same strategy, but to use Tn917 sequences as homology for integrating genomic DNA fragments into the prophage of an SPfl derivative that contained a Tn917 insertion. To provide a direct

selection for these integration events, it was necessary to place a cut gene adjacent to Tn917 sequences that were ligated to the genomic fragments. To facilitate subcloning of inserts from recombinant phage DNA it was also necessary for these Tn917 sequences’to include portions of pBR322, as explained below. A derivative of Tn917 with the required characteristics was obtained as described in Fig. 1. In this construction, approximately 3 kb of DNA internal to Tn917 has been deleted and replaced with a 1.3-kb fragment containing the cut gene from pC194 (Horinouchi and Weisbhun, 1982) and a 4.0-kb portion of pBR322 (Bolivar et al., 1977) including its rep functions and its blu gene. Approximately 1 kb of Tn92 7 DNA remains at each end of the transposon. A unique BamHI site is present within the pBR sequences, which was intended to serve as a site for inserting Sau3A-generated partial-digest fragments of genomic DNA. In addition, Sun sites flank this BamHI site. Because SuZI sites are relatively rare in SP#I DNA (Fink and Zahler, 1982), the presence of

218 /

Hind III

.Hind III

Hpa

I/w Sal I Sal11 Barn HI

Sa’I

‘Sam HI

Fig. 1. Construction of shuttle vector pTV56. Plasmids pBR322 (Bolivar et al., 1977) and pTV13 (H.-P. Biemamr, unpublished) were digested with BumHI + HindIII, ligated together and used to transform E. coliHBlOl with a selection for ApR. TcS transformants were screened for ones that contained plasmids with the structure indicated for pTV56. bla, the fi-lactamase (ApR) gene of pBR322; ret, the ‘Tc-resistance gene of pBR322; pBRrep, a region of pBR322 containing its replication functions; car, Cm acetyltransferase gene originally from pC194 (Iordanescu, 1976); Tn, sequences from Tn917 (Tomich et al., 1980); Tn(erm), the region of Tn917 containing its erythromycin-resistance gene; pE194rep, a region of pTV13 derived originally from pE194 (Iordanescu, 1976) containing replication functions active in B. subtilis.

Sal1 sites immediately flanking the primary cloning site was expected to facilitate characterization of cloned inserts carried by recombinant SP/? phages. The vector carrying this derivative of Tn917, pTV56 (Fig. l), retains pE194 rep functions external to the transposon, and can therefore replicate in B. subtilis as well as in E. coli.

digested with MI, ligated at low DNA concentration and transformed into E. coli strain HBlOl with a selection for Ap resistance (ApR). Because no SstI sites are present within pBR322 or Tn917, this resulted in a plasmid (pCV1) 11.5 kb in size that had acquired l-2 kb of SPfl DNA extending outwards from each SP/?-Tn917 insertion junction (Fig. 2).

(b) Construction of cloning vector pCV1

(c) Use of pCV1 to obtain genomic libraries

To increase homology available for recombinational transfer of cloned inserts into the SPfi prophage and at the same time eliminate pE194 rep functions, vector pCV1 was derived from pTV56 in a sequence of steps outlined in Fig. 2. First, pTV56 was linearized by digestion with HpaI and used to transform a lysogen (CU2530) containing the SPfic2A2 : : Tn917 prophage. A selection for CmR resulted in the integration into SPBc2A2 : : Tn917 of all sequences between the ends of the transposon sequences in pTV56. To obtain pCV1, a crude preparation of SP/?c2A2 : : Tn917 : : pTV56 DNA was

Digestion of pCV1 with both BamHI and SsrI would generate two DNA fragments containing, respectively, the left and right arms of Tn917 together with the additional SPfl sequences. We reasoned that if these two vector arms were ligated at high DNA concentration in the presence of DNA fragments generated by partial Sau3A digestion of the B. subtilis chromosome, intermolecular ligation events would frequently join molecules in the order: left arm chromosomal fragment - right arm. When this occurred, the left and right arms could serve as homology to integrate the chromosomal fragment

219

X

X

_-_-

____-

Tn917

select

I

c

prophage

CmR

Barn HI

Sst I ____P

SPj3 c,2 de12 :: Tn917

Sst I

I H cat m

b1aJ-j

a

I

-e-w

Sst I dilute ligation select ApR in E -. coli -

1

Fig. 2. Derivation of pCV1. Plaamid pTV56 was digested with HpaI and used to transform B. subfilti strain CU2530 (an SP/WA2 : : Tn917 lysogen) with a selection for CmR. Transformants could only arise from a double crossover involving Tn917 sequences present in the large &x11 fragment of pTV56 and their counterparts in the SPwA2: : T&l 7 prophage of CU2530. As the result of this recombination event, all sequences between the two arms of Tn.917 in pTV56 became integrated into the prophage. To obtain pCV1, phage DNA was prepared, digested with &I, ligated at dilute DNA concentration (< 1&ml), and transformed into HBlOl with a selection for ApR. pBR322-derived sequences that contain the ColEl replication origin are represented by the hatched bar; all other features of DNA fragments or plasmids are as indicated in the legend to Fig. 1.

into an SP/3: : Tn917 prophage, when the ligation mixture was used to transform an appropriate lysogen with a selection for CmR (Fig. 3). Not all CmR transformants would contain inserts, of course. For example, this would occur whenever the left arm simply became joined to the right arm at their BumHI-generated ends without an intervening chromosomal insert. However, even if this occurred ten times as often as an event that included an insert, a completely representative genomic library might be obtained in an experiment that generated only about lo4 transformants, assuming inserts to be in the size range of 5-10 kb.

To test this idea, two genomic libraries were constructed, using size-selected chromosomal DNA fragments generated by partial digestion with Sau3A, fractionated on agarose gels and recovered by electroelution. One of these libraries (library A) was constructed with fragments in the size range of 5-8 kb. The other (library B) was constructed with fragments in the size range of 8-10 kb. In both cases, about 10 pg of size-selected fragments were ligated in the presence of roughly equimolar amounts of the vector arms. The ligations were carried out at high DNA concentration (>50 pg/ml) to ensure that

220

I

*left

arm pCVl+

I t

partial Sau 3A right arm pCVl-

chromosome -

I

I--

I

I sit

sit

I

I

ligate transform

Tn917

J

select

SPB c2

de12 :: TnQ 17 lysogen

SPB CA de12 :: Tn917

prophage

CmR Eco RI DBR

I

--recombinant

prophage

Sai T Fig. 3. Use of pCV1 to obtain a genomic library by prophage transformation. pCV1 was digested with both BumHI and SsrI to produce the two vector arms. These were mixed in a 1: 1 molar ratio with size-selected chromosomal DNA fragments generated by partial digestion with Suu3A and ligated at high DNA concentration (> 50 r&ml) to favor the formation of long concatemers through end-to-end joining ofditferent molecules. Whenever these ligation events joined fragments in the sequence: letI vector arm - chromosomal fragment - right vector arm, the cut gene from the left arm + the chromosomal fragment + pBR sequence, as a unit, became flanked by SPfl and Tn9f 7 sequences that provided 3-4 kb of homology on either side to transfer this unit by recombination into an SPb2d2 : : Tn.917 prophage, as shown schematically in the figure.

intermolecular joining events were favored. Ligation mixtures were used to transform B. subtih strain PY480 (an SPfic2d2 : : Tn917 lysogen) with a selection for CmR. In the case of library A, 1.2 x lo5 transformants were obtained, and in the case of library B, 8 x 104 transformants were obtained. Although no effort was made to determine the actual percentage of transformants that included cloned inserts, we estimate that the libraries should have been completely representative of genomic sequences even if this frequency were as low as 5 %.

(d) Screening of the libraries for cloned inserts To recover specialized transducing phages that could be tested for complementation of specific mutations, CmR transformants from each genomic library were pooled (library A and library B transformants were pooled separately), and these pooled mixtures were induced to produce phage lysates. Because the SP/3 prophage carried the c2 mutation, which makes the phage repressor thermolabile, lysates were produced with a simple heat-induction

221

procedure (Rosenthal et al., 1979). These library lysates were in turn used to infect SPp strains containing various different auxotrophic mutations (CU1065, CU1441, CU1442, CU1443 or CU1448), and selections were made simultaneously for CmR and specific prototrophic phenotypes. Attempts were made to complement nine different auxotrophic cysA 14, &Z-l, mutations, including a&906, glyB 133, metC3, purA 16, purB33, thrA5 and trpC2, which represent widely scattered chromosomal regions. Although relative numbers of CmR and prototrophic transductants varied considerably for the different markers used (data not shown), complementation was observed in all cases. Transductants from each selection were pooled, heat-induced again, and again used in selections for the same prototrophic phenotypes. This was done to enrich for the more stable and replication-proficient recombinant phages. Individual transductants obtained in the second round of selections were colony-puritied and characterized in more detail to determine whether they actually contained non-defective recombinant phages that transduced the complementing phenotypes. First, transductants were tested to determine whether loss of the prophage restored an auxotrophic phenotype. To cure of the prophage, lysogens were streaked for single colonies on non-selective media at 50°C (Rosenthal et al., 1979). Transductants that lost the complementing phenotype when they were cured of the prophage were tested to determine whether, when subjected to heat-induction, they produced a lysate that transduced CmR together with prototrophy at high frequency. Lysates were used to infect auxotrophic strains and selections were fast made for CmR only on complete media. CmR transductants were then tested for complementation of the relevant auxotrophic mutations. For all nine autotrophic mutations examined, non-defective recombinant phages were obtained that transduced both CmR and prototrophy at high frequency. Linkage of the two phenotypes was usually less than 100 % (Table II), however, indicating some tendency to delete cloned inserts during phage propagation. In addition to the nine auxotrophic mutations examined, we also selected for the recovery of recombinant phages that could complement the recE4 mutation ofB. subtibk Like red mutations in E. coli, the recE4 mutation in B. subtilti blocks homologous

TABLE II Structural stability of recombinant phages Mutation complemented*

% Stabilityb Library A

aroI906

88 (44/50)

cysA 14

92 (46/50)

dal-1 glyB133 metC3 purA 16 purB33 thrA 5 trpC2 recE4

96 (48/50)

100 (SO/SO) 64 (32/50) 52 (26/50)

N.D. N.D. N.D. 96 (48150)

Library B N.D. 92 (46/50) 100 (50/50) N.D. 56 (28/50) 20 (10/50) 96 (48/50) 94 (47/50) 80 (40/50) N.D.

a Mutant alleles were carried by strains CU1065, CU1441, CU1442, CU1443, CU1448 or YB1015 (see Table I). b Stability was determined by infecting the relevant mutant with a lysate of the recombinant phage followed by selection for CmR on LB. Transductants were then patched onto appropriately supplemented TSS medium (in the case ofthe auxotrophs) or LB containing 0.04 pg MC/ml (in the case of YBlOl5) to score for complementation. The numbers in parentheses indicate the actual numbers of transductants scored and the numbers that retained the complementing phenotypes. N.D., not determined.

recombination and confers sensitivity to DNAdamaging agents such as EMS and MC (Dubnau et al., 1973). Thus we could screen lysates from library A or library B for phages that could transduce a recE4 mutant (YBlOl5) simultaneously to CmR and MCR. Nondefective specialized transducing phages of this type were recovered from both genomic libraries. (e) Attempts to transfer cloned sequences from phages to plasmids by homologous recombination The general structure of recombinant phages obtained as described above is such that the cloned inserts are always flanked on one side by a cat gene and on the other side by pBR sequences (Fig. 3). Moreover, the cat gene + insert + pBR sequences, as a unit, are flanked by the two arms of Tn9I7. Thus, in principle, it should be possible to transfer this unit by homologous recombination, selecting for CmR, to any B. subtilis plasmid that contains a copy of Tn92 7 (Fig. 4). Because recombinational transfer of the cloned insert would be accompanied by transfer of pBR sequences as well, the result would be to

222

trarwforming

fragment

of

SpB c,2 de12 :: Tng 17 :: cloned recipient

inaert

plasmid

relect

CmR

insert-containing shuttle

replicon

0 0 I . 0”

Fig. 4. ‘Recombinational subcloning’ of inserts from SPj?c2d2::Tn917 to pTV5. In SPjk2A2::Tn917 specialized transducing phages containing cloned inserts, a cat gene selectable in B. subtih + the insert + pBR sequences, as a unit, are flanked by DNA sequences including the two arms of Tn927. As indicated schematically in the figure, the Tn917 arms provide homology to transfer this unit by homologous recombination to a Tn92 7-containing plasmid like pTV5. This is accomplished by preparing a small quantity of phage DNA and using it to transform competent B. subrilis ceils that already contain pTV5, selecting for Cm resistance. The result is an insertcontaining biinctional ‘shuttle replicon’ that can replicate in both E. coli and B. subtih.

kb

Fig. 5. Restriction digests of recombinant phages and insert-containing shuttle replicons derived from them as described in Fig. 4, fractionated by electrophoresis through 0.7% agarose. Hind111 digests of wt coliphage 1 DNA were run in lanes 1 and 11 as size standards, with fragment molecular weights indicated at the left. All other lanes contain WI digests: 2, SP/?c2AZ::Tn917; 3, SPwA2 : :Tn9Z 7: : pTV56; 4, a recombinant phage obtained from library B that complements the U-1 mutation; 5, an insert-containing shuttle replicon derived from that phage; 6, a recombinant phage obtained from library A that complements the da/-1 mutation; 7, an insert-containing shuttle replicon derived from that phage; 8, a recombinant phage obtained from library B that complements the trpC2 mutation; 9, an insert-containing shuttle replicon derived from that phage; 10, pTV56. Arrows mark the positions of Sal1 fragments in digests of recombinant phages that corn&rate with insert-containing Sal1 fragments in shuttle replicons derived from those phages.

223

produce a shuttle plasmid that could replicate in E. coli, significantly facilitating further analysis of the

inserts. To determine how well such ‘recombinational subcloning’ might work in practice, a small quantity of DNA was prepared from recombinant phages that complemented the aro1906, cysA 14, dal-1, glyBI33, purB33, thrA5, trpC2 and recE4 mutations, and these DNA samples were used to transform an SP/Is B. subtilis strain (PY468) that contained pTV5 (Youngman et al., 1984) with a selection for CmR. Transformants were readily obtained in all cases, and, for each phage DNA sample, at least some of the transformants contained plasmid DNA that could be transformed into E. cob strain HBlOl to yield ApR transformants. In only two cases, however, involving DNA samples from recombinant phages that complemented the dal- 1 and trpC2 mutations, did these ApR transformants retain a cloned insert. In these two cases, the sizes of inserts transferred to pTV5 were apparently identical to the sizes of inserts present in the original recombinant phages

(Fig. 5). Moreover, E. coli plasmids containing these inserts could transform &Z-l and rrpC2 mutants to prototrophy. In all other cases, ApR transformants contained a plasmid similar in size to pTV56 (data not shown), indicating that the entire insert had been deleted. Recently, in tracing the subcloning history of the cut fragment inserted into Tn917 to produce pTU13 (Fig. l), we determined that this fragment must include approximately 275 bp of DNA derived from pBR322 (unpublished data). As a consequence, in the construction of pTV56 an interval of 275 bp of pBR DNA was duplicated in a direct-repeat orientation, on either side of the BamHI site. Recombination between these 275-bp repeats could have been responsible for deleting cloned inserts transferred to pTV5 by recombinational subcloning. (f) ‘Rescue’ of cloned material directly from phage DNA digests

Because cloned inserts are immediately adjacent to pBR sequences in the recombinant phages (Fig. 3,

a E $-x$

s

s

’

m

:

b

C

d

Fig. 6. Linear physical maps ofplasmids derived from recombinant phages by ‘rescue’intoE. coli of pBRrep-containing EcoRI fragments. Only /&RI and Sal1 sites are shown. To the right ofthe maps: a, pCV1; b, a plasmid derived from SP~AZ::Tn917::pTV56 (no insert); c, a plasmid derived from a recombinant phage that complemented purB33; d, a plasmid derived from a recombinant phage that complemented aro1906; e, a plasmid derived from a recombinant phage that complemented cyrA 14; f, a plasmid derived from a recombinant phage that complemented &AS; g, a plasmid derived from a recombinant phage that complemented recE4. Above the maps: E, EcoRI sites; S, Sal1 sites. The solid bars indicate sequences presumed to represent residual portions of cloned inserts. For hatched bars, see Fig. 2.

224

bottom), we expected that it might often be possible to recover all or parts of the inserts simply by cutting phage DNA with various restriction enzymes, circularizing digest fragments by dilute ligation, and transforming into an E. coli strain with a selection for ApR. For example, if an insert contained no EcoRI site, an EcoRI digest of the recombinant phage would generate an insert-containing fragment that also contained not only pBR rep functions but the cat and blu genes as well (Fig. 3). In such a case, circularized fragments recovered in E. coli would include the entire insert. If the insert contained one or more EcoRI site(s), recovered fragments would include only part of the insert and would lack the cut gene. To evaluate this approach, we prepared an EcoRI limit digest of live of the recombinant phages from which we had failed to recover a cloned insert by recombinational transfer to pTV5 (phages complementing the purB33, aroZ906, cysA 14, &A 5 and recE4 mutations). These digests were ligated at low DNA concentrations (< 1 pg/ml) and used to transform E. coli HB 101. In all cases, ApR transformants were readily obtained. In four of the five cases, phages complementing purB33, aroZ906, cysA 14 and thr,45, these transformants contained plasmids whose structures were consistent with their having been derived from recombinant phages containing inserts (Fig. 6). This tends to support the conclusion that inserts are relatively stable in the recombinant phages and are probably deleted after recombinational transfer to pTV5, as suggested above. In one case (a phage complementing recE4), six of six transformants contained plasmids whose structures indicated a deletion of the entire insert (Fig. 6). This may be an exceptional case, reflecting the fact that the recE gene is difficult to propagate on multicopy vectors (R.E. Yasbin, personal communication). In the four cases where plasmids appeared to retain a portion of the cloned insert, plasmid samples were transformed back into B. sub& strains selecting for recombinational correction of the mutant alleles complemented by the original recombinant phages. Only in the case of plasmids derived from a phage complementing the purB33 mutation were prototrophic transformants obtained. These plasmids apparently contained a 3.25-kb portion of the cloned insert (Fig. 6).

(g) Conclusions and considerations for vector improvement We have demonstrated the successful use of a new kind of prophage transformation cloning system for B. subtilis. Important advantages of this system include: (1) the ease with which a broadly representative genomic library can be produced; (2) the fact that non-defective recombinant phages can be recovered that contain inserts up to at least 10 kb in size, and (3) the possibility in at least some cases of subcloning inserts from phages onto plasmids either by recombinational transfer or by rescuing portions of inserts into E. coli on restriction fragments from the phage that include pBR rep functions. Because no other cloning strategy offers such features, we consider this system to represent a highly promising approach for the cloning of B. subtilis DNA in a B. subtilis host. Nevertheless, the system in its present form has not met all of our expectations, and we consider below several ways of improving its various components. Our relatively low success rate in transferring inserts from phages to plasmids by homologous recombination, and in stably maintaining them in a plasmid-borne state, seems most likely to reflect flaws in the specific vectors we used rather than a general problem with the strategy itself. First, the BamHI cloning site in pCV1 was flanked by duplicated sequences, which probably facilitated the deletion of inserts, after transfer to pTV5, through intramolecular recombination. Second, pTV5 was probably a poor choice as a target for recombinational subcloning, because it depends for replication on pE194-derived rep functions and thus produces significant quantities of single-stranded DNA (te Riele et al., 1986). This would probably increase the likelihood of insert deletion even in the absence of obvious direct repeats flanking the cloning site (S .D. Ehrlich, personal communication). For this reason, we will construct for future work a new transfer target consisting of a Tn9Z 7-containing derivative of pAMfl1 (Clewell et al., 1974), which is known to produce little or no single-stranded DNA during replication (L. Jarmike and S.D. Ehrlich, personal communication). Finally, we emphasize the ease with which we were able to rescue portions of cloned inserts by digestion of recombinant phages with EcoRI, fol-

225

lowed by circularization of fragments and transformation of an E. cofi recipient. The most important limitation to this approach is the fact that inserts as large as 8-10 kb would normally contain one or more EcoRI site. It occurs to us that ifthe two EcoRI sites in pCV1 were converted to Not1 sites, which are very rare in B. subtilis DNA, this approach might in general be the method of choice for recovering cloned inserts from phages.

ACKNOWLEDGEMENTS

We are grateful to S.A. Zahler and members of his group at Cornell for gifts of strains and for many useful discussions. We thank H.-P. Biemann for the construction of pTV13 and J. Westpheling for the construction of pTV56. This work was supported by NIH grant GM35495 and a postdoctoral fellowship to H.P. from the Deutscher Akademischer Austauschdienst Sonderprogramm Gentechnologie.

REFERENCES Anagnostopoulos, C. and Spizizen, J.: Requirements for transformation in Badussubtibk J. Bacterial. 81 (1961) 741-746. Bimboim, H.C. and Daly, J.: A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7 (1979) 1513-23. Bolivar, F., Rodriguez, R.L., Greene, P.J., Betlach, M.C., Heynecker, H.L., Boyer, H.W., Crosa, J.H. and Falkow, S.: Construction and characterization of new cloning vehicles, II. A multipurpose cloning system. Gene 2 (1977) 95-113. Bonamy, C. and Szulmajster, J.: Cloning and expression of Bncillus subtifis sporulation genes. Mol. Gen. Genet. 188 (1982) 202-210. Boyer, H.W. and Roulland-Dussoix, D.: A complementation analysis of the restriction and modification of DNA in Escherichia coli. J. Mol. Biol. 41 (1969) 459-472. Clewell, D.B., Yagi, Y., Dunny, G.M. and Schultz, SK.: Characterization of three plasmid deoxyribonucleic acid molecules in a strain of Streptococcus faecalis: Identification of a plasmid-determined erythromycin resistance. J. Bacterial. 117 (1974) 283-289. Cohen, S.N., Chang, A.C.Y. and Hsu, L.: Nonchromosomal antibiotic resistance in bacteria: genetic transformation of Escherichiu coli by R-factor DNA. Proc. Natl. Acad. Sci. USA 69 (1973) 21 lo-21 14. Dubnau, D.: Molecular cloning in Bacilfus subtibk In Inouye, M. (Ed.), Experimental Manipulation of Gene Expression, Academic Press, NY, 1983, pp. 33-51.

Dubnau, D., Davidoff-Abelson, R., Scher, B. and Ciigliano, C.: Fate of transforming deoxyribonucleic acid after uptake by competent Bacillus subtilir: phenotypic characterization of radiation-sensitive recombination-deficient mutants. J. Bacteriol. 114 (1973) 273-286. Ehrlich, S.D., te Riele, H., Petit, M.A., Jam&i, L., Noirot, P. and Michel, B.: DNA recombination in plasmids and the chromosome of Bacillus subtilk. In Ganesan, A.T. and Hoch, J.A. (Eds.), Bacillus Molecular Genetics and Biotechnological Applications. Academic Press, New York, 1986, pp. 27-34. Errington, J.: Generalized cloning vectors for Bacillus sub&s. In Denhardt, D.T. and Rodriguez, R.L. (Eds.), Vectors: A Survey of Molecular Cloning Vectors and Their Uses. Butterworth, Boston, MA, 1987, pp. 345-362. Errington, J. and Jones, D.: Cloning in BociZlussubtilis by transfection with bacteriophage vector $105527: isolation and preliminary characterization of transducing phages for 23 sporulation loci. J. Gen. Microbial. 133 (1987) 493-502. Fink, P.S. and Zahler, S.A.: Restriction fragment maps of the genome ofBaci&ssubtiZis bacteriophage SPfl Gene 19 (1982) 235-238. Guzman, P., Westpheling, J. and Youngman, P.: Characterization of the promoter region of the Bacillus subtiks spoZZE operon. J. Bacterial. 170 (1988) 1598-1609. Hahn, J. and Dubnau, D.: Analysis ofplasmid deletion instability in BaciZZussubtdis. J. Bacterial. 162 (1985) 1014-1023. Horinouchi, S. and Weisblm, B.: Nucleotide sequence and functional map of pC194, a plasmid that specifies inducible chloramphenicol resistance. J. Bacterial. 150 (1982) 815-825. Iordanescu, S.: Three distinct plasmids originating in the same Staphylococcus aureus strain. Arch. Roum. Path. Exp. Microbiol. 35 (1976) 125-159. Jones, D. and Errington, J.: Construction of improved $105 vectors for cloning by transfection in Baci2Zu.ssubtilis. J. Gen. Microbial. 133 (1987) 483-492. Kawamura, F., Saito, H. and Ikeda, Y.: A method for construction of specialized transducing phage pl 1 of Bacillus sub&s. Gene 5 (1979) 87-91. Rosenthal, R., Toye, R.A., Korman, R.Z. and Zahler, S.A.: The prophage of SP&?dcitK-I, a definitive specialized transducing phage of Bacillw subtibk. Genetics 92 (1979) 721-739. Saava, D. and Mandelstam, J.: Cloning of the Bacillus sub&s spoZZA and spoVA loci in phage $lOSDI: It. J. Gen. Microbiol. 130 (1984) 2137-2145. Segall, J. and Losick, R.: Cloned Bacillus subtilisDNA containing a gene that is activated early during sporulation. Cell 11 (1977) 751-761. Seki, T., Miyaki, H., Yoshikawa, H., Kawamura, F. and Saito, H.: An improved method of prophage transformation in Bacillus subtilis. J. Gene. Appl. Microbial. 32 (1986) 73-79. te Riele, H., Michel, B. and Ehrlich, S.D.: Are single-stranded circles intermediates in plasmid DNA replication? EMBO J. 5 (1986) 631-637. Tomich, P., An, F.Y. and Clewell, D.B.: Properties of erythromycin inducible transposon TrQ17 in Streptococcus faecaks. J. Bacterial. 141 (1980) 1366-1374. Youngman, P.: Plasmid vectors for recovering and exploiting

226 Tn917 transpositions in Bacillus and other Gram-positives. In Hardy, K. (Ed.), Plasmids: A Practical Approach. IRL Press, Oxford, U.K., 1987, pp. 79-103. Youngman, P., Perkins, J.B. and Losick, R.: Genetic transposition and insertional mutagenesis in Bacillus subtilis with Streptococcus faecah transposon Tn917. Proc. Natl. Acad. Sci. USA 80 (1983) 2305-2309. Youngman, P., Perkins, J.B. and Losick, R.: Construction of a cloning site near one end of Tn917 into which foreign DNA may be inserted without affecting transposition in Bacillus

subtih or expression

of the transposon-borne erm gene. Plasmid 12 (1984) 1-9. Zahler, S.A., Korman, R.Z., Odebralski, J.M., Fink, P.S., Mackey, C.J., Pout& C.G., Lipsky, R. and Youngman, P.: Genetic manipulations with phage SPfi. In Ganesan, A.T., Chang, S. and Hoch J.A. (Eds.), Molecular Cloning and Gene Regulation in Bacilli. Academic Press, San Diego, CA, 1982, pp. 41-50. Communicated by J.A. Hoch.