University of Groningen

Replication and maintenance of plasmids in Bacillus subtilisMeijer, Wilhelmus Johannes Jozef

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:1995

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Meijer, W. J. J. (1995). Replication and maintenance of plasmids in Bacillus subtilis. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 09-10-2020

59

Chapter III

Effects of the generation of single-stranded DNA on the maintenance ofplasmid pMV158 and derivatives in different Bacillus subtilis strains

Wilfried J. J. Meijer, Daniël van der Lelie, Gerard Venema and Sierd Bron

Published in: Plasmid (1995) 33; 79-89

Chapter III

60

SUMMARY

The effects of the single-strand origins (SSOs) of plasmid pMV158 on (i)the conversion of its single-stranded (ss) replication intermediates t odouble -stranded (ds) plasmid DNA and (ii) its maintenance, wer eanalyzed. The roll ing-circle plasmid pMV158, which replicates via ssDNAintermediates, contains two single-strand origins (SSOs) of replication ,palA and palU. In this chapter the results obtained with Bacillus subtilisare described; complementary studies with Lactococcus lactis arepresented in chapter II. While in L.lactis both SSOs are functional a sssDNA conversion signal, only palU appeared to be active B.subtilis.Similar to the situation in L.lactis, the accumulation of large amounts ofssDNA resulted in a severe decrease in plasmid maintenance in B.subtilis.In the latter bacterium large amounts of ssDNA were only accumulated,however, when plasmids lacking a functional SSO were propagated i nRecA mutant strains. In wild type RecA strains these plasmid saccumulated only modest amounts of ssDNA and they were maintainedat fairly stable levels. The results suggest that in B.subtilis a RecA-mediated alternative pathway exists for the conversion of ssDNA whichcan improve plasmid maintenance. In addition to ssDNA accumulation,and the antagonizing role of RecA therein, two other plasmi regions wereshown to affect pMV158 maintenance in B.subtilis. One was mob generegion, which had a negative effect on plasmid maintenance; and th eother the palA type SSO. Although palA was not functional as a ssDNAconversion signal in B.subtilis, its presence had a positive effect o npMV158 maintenance.

INTRODUCTION

Most small multicopy plasmids fromGram-positive bacteria replicate via therolling-circle mechanism (RCM; forreviews see (Gruss and Ehrlich, 1989;Novick, 1989; Jannière et al., 1993), whichgenerates single-stranded plasmid DNA(ssDNA) intermediates. For the efficientinitiation of ssDNA conversion to double-stranded (ds) plasmid DNA, single-strandorigins (SSOs; formerly referred to asminus origins) are required. Most SSOs thathave been identified so far depend on thehost-encoded RNA-polymerase to primecomplementary DNA strand synthesis on

the ssDNA template (Gruss et al., 1987;Boe et al., 1989; Devine et al., 1989; Bron,1990).

Owing to the fortuitous expressionof their natural antibiotic resistancemarkers in a wide variety of bacteria, manyof the cloning vectors used today for Gram-positive bacteria are based on RCMplasmids derived from Staphylococcusaureus. Since SSOs of RCM plasmids areusually only functional in their native host,inefficient ssDNA conversion may occurwhen vectors based on these plasmids areused in non-native hosts (Gruss et al., 1987;Del Solar et al., 1987; Bron, 1990). Thisresults in the intracellular accumulation ofssDNA intermediates. ssDNA accumulationis frequently associated with a decrease in

Chapter III

61

Figure 1. Linear maps of pMV158 and pUB110 derivatives.(A): pMV158 and derivatives. (B): Derivatives of pUB110 (high- and low-copy number variants )containing palA of pMV158. The positions of palU, the mob gene (Mob) and its promoter (p), palA, thedouble-stran d origin (ori), the replication regions, (repA/repB in pMV158 and repU in pUB110), and thetetracy cline, the kanamycin and bleomycin resistance genes (Tet/Tc , Km , and Bm ) are indicated. OnlyR R R

restriction sites relevant for constructions are indicated.

plasmid maintenance (Bron et al., 1988a; subtilis are described. ComplementaryBron et al., 1991; Viret and Alonso, 1988). studies with Lactococcus lactis are

The aim of the present presented in chapter II.investigations was to analyze possiblecauses of poor maintenance of RCMplasmids in Gram-positive bacteria. Inparticular, we addressed the questionwhether ssDNA accumulation by the broad-host-range RCM plasmid pMV158 resultsin decreased plasmid maintenance. pMV158was originally obtained from Streptococcusagalactiae (Burdett, 1980). In the presentchapter the results obtained with Bacillus

RESULTS

Construction of pMV158 derivatives To study the effects of the SSOs and

the mob gene of pMV158 on ssDNAaccumulation and plasmid maintenance inB.subtilis, several deletion derivatives ofpMV158 were constructed (Figs. 1A and1B). In the nomenclature used for thederivatives the letters "U" and "A"

Chapter III

62

symbolize the palU and palA-type SSO. The Since the copy numbers of thesepresence or absence of an SSO is indicated pUB110 derivatives are about 3- to 5-foldby U (or A ) and U (or A ), respectively. higher than those of pMV158, we also+ + - -

Plasmid pMV(U A ) is identical to the constructed low-copy variants of pWUB10- +

formerly described pLS1 (Lacks et al., and pWUB20. This was accomplished by1986). Details of the construction exchanging the 3700 bp AccI/BamHIprocedures for pMV(U A ), pMV(U A ), fragment of pWUB10 and pWUB20 by the- + + -

pMV(UA ), pMV(mob ) and pMVCm (Fig. corresponding fragment from pUB110cop1- - *

1A) are described in chapter II. In pMV(U (Leonhardt, 1990). The resulting variants,-

A ) the palU SSO and the 5'-part of the mob pWUB11 and pWUB21, have copy numbers+

gene are absent and in pMV(U A ) the palA of about 5 to 10 per chromosome+ -

SSO and the 3'-part of the mob gene were equivalent. pMV(U A )Em is a spontaneousremoved. pMV(U A ) lacks both SSOs and high-copy-number variant of pMV(UA ) in- -

the entire mob gene. pMV(mob ) is the which the tetracycline resistance (Tc ) gene*

derivative in which the pMV158-located was replaced by the erythromycinmob gene, required for conjugative resistance (Em ) gene of pE194. The Tcmobilization of the plasmid (Priebe and gene was replaced because it was expectedLacks, 1989; van der Lelie et al., 1990), was that large amounts of the Tc gene product,inactivated by a frameshift mutation. In synthesized from the high-copy plasmid,pMVCm, the mob gene was replaced by the might be toxic to the cells (de la Campa etchloramphenicol-resistance (Cm ) gene of al., 1990). Sequence-analysis of the high-R

Bacillus pumilus. copy-number mutation revealed thepWUB10 and pWUB20 (fig. 1B) are addition of one A-residue at position 658 of

derivatives of the RCM plasmid pUB110 in pMV(U A ) [coordinates according towhich the palA SSO of pMV158 was (Lacks et al., 1986)]. As a consequence, theintroduced. pUB110 naturally contains palU gene specifying the RepA protein, which isand a mob gene related to that of pMV158 a repressor for the control of replication(van der Lelie et al., 1989). PalA of (Del Solar et al., 1990; Pérez-Martin andpMV158 was cloned at the 3'-end of the Espinosa, 1991) is out of frame after themob gene of pUB110 to mimic the situation third codon. This resulted in a copy numberin pMV158. To that aim, pMV158 was of about 100 per chromosome equivalent.digested with Csp45I and XhoI, andpUB110 was linearized with EcoRI. The 5'ends of the purified 813 bp fragment ofpMV158, containing palA, and the linearpUB110 were filled-in using Klenow DNApolymerase. After ligation, the mixture wasused to transform B.subtilis PSL1protoplasts. Restriction analysis revealedthe presence of palA-containingrecombinant plasmids of the expectedstructure in kanamycin-resistant cells. ThepUB110 derivative containing palA in thesame orientation relative to the double-strand origin of replication as in pMV158(the functional orientation) was designatedpWUB10, and the derivative with palA inthe reversed orientation pWUB20.

- -

- -

R

R R

R

- +

Effect of the single-strand origins an dthe Mob protein on plasmid cop ynumbers

The copy numbers of the differentconstructs were determined from the ratioof radiolabeled plasmid versuschromosomal DNA. The averages of fourindependent experiments are listed in Table1. Except for the high-copy-number mutantpMV(U A )Em (about 100 copies per- -

chromosome equivalent), all pMV158derivatives had roughly similar copynumbers in B.subtilis (from about 10 to 17per chromosome equivalent). The copynumbers of pMV158 variants lacking one orboth SSOs were not detectably reduced in

Chapter III

63

Figure 2. Total DNA lysates of B.subtilis strains 8G5 (A) and 7G224 (B)harbouring various plasmids. Samples were in cubated without (-), or with (+) S1-endonuclease, or werelinearized (li); m : EcoRI digested SPP1 phage DNA used as molecularweight reference marker. The fragments have the following sizes (in kb):8.5; 7.4; 6.15; 4.95; 3.59; 2.84; 1.97; 1.91, 1.56; 1.45; 1.16; 0.98; 0.72.The position of the various plasmid DNA forms are indicated: ssDNA (ss),supercoiled (sc), linear (li) and open circular (oc).

comparison to pMV158. RecA protein (de Vos and Venema, 1981).

Effect of the single-strand origins an dthe mob gene on the accumulation of ssplasmid DNA in different B.subtili sbackgrounds

The effects of palU and/or palA, andof the mob gene on the efficiency ofconversion of ss to ds plasmid DNA wereexamined in the B.subtilis strains 8G5 and7G224. Apart from the recA4 allele, strain7G224 is isogenic to strain 8G5 whichcontains the wild-type recA allele (de Vosand Venema, 1981). The recA4 mutation isa point mutation resultin in an inactive

Total DNA extracts from B.subtilis 8G5cells harbouring pMV158 or its derivativeswere divided in two portions, one of whichwas treated with S1-endonuclease to digestssDNA; the other portion was leftuntreated. The results of Southernhybridizations of blots from agarose gelsare shown in Fig. 2A. With pMV158,containing both SSOs, no ssDNA could bedetected indicating that the conversion ofssDNA was efficient. Inactivation of themob gene by a frameshift mutation[pMV(mob )] or the replacement of the*

entire mob gene by the Cm gene did notR

Chapter III

64

Figure 3. Maintenance of pMV158 and derivatives in the B.subtilis strains 8G5 (A) and 7G224(B).# : pMV158; + : pMV(U A ); è : pMV(U A ); > : pMV(U A ). - + + - - -

result in detectable levels of ssDNA accumulation was observed with pMV(UA ) and pMV(U A ) when the DNA extracts

results indicated that no structural or were prepared from the recA4 strain 7G224functional elements from the mob gene (Fig. 2B). No differences in ssDNAregion were required for efficient ssDNA accumulation between the two strains wereconversion. observed with constructs containing a

When only palU was present functional SSO (i.e. palU; results not[pMV(U A )], no ssDNA was observed shown). These results indicate that RecA+ -

under the conditions used. Some ssDNA can assist in the conversion of ssDNA tocould be detected, however, when the gels dsDNA of pMV158 derivatives devoid ofwere overloaded (results not shown). With functional SSOs. pMV(UA ) ssDNA was observed. Deletion- +

of both SSOs and the mob gene [pMV(UA- -

)], resulted in about the same level ofssDNA accumulation as when only palAwas present. These results indicated that,whereas palU is efficient in B.subtilis, palAhas no, or only little activity in thisorganism. Remarkably, even without afunctional ssDNA conversion signal [i.e.pMV(U A ) and pMV(U A )], only modest- + - -

amounts of ssDNA were detected in strain8G5, indicating that in these cases ssDNAconversion still occurred rather efficient.Strikingly, a large increase in ssDNA

-

accumulation (results not shown). These + - -

Effects of ssDNA accumulation o nplasmid maintenance

The possible effects of ssDNAaccumulation on the maintenance of thevarious pMV158 derivatives were measuredin B.subtilis 8G5. The results are presentedin Fig. 3A. The derivatives containingeither palA or palU were maintained quitestably (more than 90% of the cellscontained the plasmid after 80 generationsof growth in non-selective media). pMV(U-

A ), lacking both SSOs, was rendered-

moderately unstable (30% of the cells

Chapter III

65

Figure 4. Maintenance of pMV(U A ) and the- -

high-copy number plasmid pMV(U A )Em in- -

B.subtilis 8G5. > : pMV(U A ); G : pMV(U A )Em- - - -

contained the plasmid after 80 generationsof non-selective growth). Unexpectedly,pMV158, containing both SSOs, was themost unstable of all plasmids tested (lessthan 5% of the cells harboured the plasmidafter 80 generations of growth). The shapeof the curve, showing that the kinetics ofloss of pMV158 was biphasic, is discussedin one of the following sections. We nextstudied whether the differences in ssDNAaccumulation between the recA wild typeand recA4 mutant strains were reflected indifferences in plasmid maintenance. Asshown in Fig. 3B, the maintenance ofpMV(U A ) and pMV(U A ) was strongly- - - +

decreased in the recA4 strain. Hence, largeamounts of ssDNA were paralleled by lowlevels of plasmid maintenance. In addition,in the wild type recA background,moderate [pMV(U A )] and low amounts- -

of ssDNA [pMV(U A )] were paralleled+ -

with a moderate and high level ofmaintenance, respectively.

Maintenance of pMV158 andderivatives was, however, also clearlyaffected by factors other than ssDNAaccumulation. This conclusion is based onthe following observations made in the8G5 recA wild type background:(a): despite the accumulation of similaramounts of ssDNA with pMV(UA ) and- +

pMV(U A ), the former plasmid was- -

maintained more stably than the latter.Interestingly, the superior maintenance ofpMV(U A ) over pMV(U A ) was most- + - -

pronounced in the recA wild typebackground (Figs. 3A and 3B).(b): whereas no ssDNA was detected withpMV158, (containing both SSOs), thisplasmid was highly unstable.

The maintenance of pMV(U A ) is not- -

improved by increasing the cop ynumber

Up to now no active partitioningfunctions have been identified on RCMplasmids. Therefore, these plasmids arelikely to be partitioned randomly and, as aconsequence, increased plasmidmaintenance is expected with increasedcopy numbers. To test this prediction, themaintenance of the high-copy-numbervariant pMV(U A )Em was compared to- -

that of pMV(U A ) in strain 8G5. Fig. 4- -

shows that the maintenance of the high-copy-number mutant pMV(UA )Em was- -

not increased relative to that of pMV(UA ).- -

In fact, the increase in copy number seemedto further decrease the stability of theplasmid.

Chapter III

66

Figure 5. Maintenance of pMV158 ( ##),pMV(mob ) (ªª), and pMVCm ( !!) in*

B.subtilis 8G5.

Sequences within the mob gene underliethe high instability of pMV158 i nB.subtilis

Fig. 3A shows that the kinetics ofplasmid loss of pMV158 was biphasic.Although the onset of the high rate of rapidplasmid loss differed between 30 and 60generations of growth in independentexperiments, the biphasic nature of thekinetics of pMV158 loss was consistentlyobserved. Biphasic segregation kinetics areindicative for the presence of mixedpopulations of cells, in which one cell typehas a growth advantage over the other (Boeet al., 1987). To test whether growthadvantages underlie the biphasic nature ofthe kinetic curves here, we measured thegrowth rates of plasmid-containing andplasmid-free cells in mixed cultures. Cellscontaining pMV158, pMV(UA ),- +

pMV(U A ) and pMV(U A ) were used in+ - - -

this test. When mixed in a 1:1 ratio withplasmid-free cells, the frequency ofplasmid-containing cells remainedapproximately constant over 20 generationsof growth with pMV(UA ), pMV(U A ) or- + + -

pMV(U A ). However, with pMV158 the- -

ratio of plasmid-free versus plasmid-containing cells increased from 1 to 18.5after 20 generations of growth. Clearly, thepresence of pMV158 caused growthretardation to the cells.

Several reasons for the highinstability of pMV158 were considered. (a) To study whether the instability wascaused by the presence of the Mob protein,the maintenance of pMV(mob), producing*

an inactive Mob protein, was examined.pMV(mob ) showed the same low stability*

as pMV158 (Fig. 5). As with pMV158, cellscontaining pMV(mob ) were rapidly*

outgrown by plasmid-free cells (results notshown). Apparently, the low stability ofpMV158 was not associated with thepresence of functional Mob protein.(b) To test whether the joint presence of

palA and palU caused the high levels ofinstability of pMV(mob ) and pMV158, the*

maintenance of pWUB10, -20, -11 and -21was tested. These plasmids are high-copynumber (pWUB10 and pWUB20) and low-copy number (pWUB11 and pWUB21)derivatives of pUB110 which carry the palUSSO (from pUB110) and the palA SSO(from pMV158) at similar positions relativeto the mob gene as pMV158. The aminoacid sequences of the Mob proteinsspecified by pMV158 and pUB110 show45% identity in the N-terminal 200 aminoacids (van der Lelie et al., 1989). Noplasmid loss was observed with any of thepWUB plasmids during 100 generations ofgrowth (results not shown). This indicatedthat the joint presence of palA and palU onthe same replicon did not cause instabilityof pUB110 derivatives. We consider it,therefore, also unlikely that the presence ofthese two SSOs caused the instability of

Chapter III

67

pMV158 and pMV(mob ) in B.subtilis 8G5. causing a decrease in the maintenance of*

(c) We next tested whether the replacement pMV158 derivatives (Meijer et al., 1995).of the entire mob gene by the Cm gene The present work showed that also inR

(pMVCm) was able to restore plasmid B.subtilis large amounts of ssDNA weremaintenance. No ssDNA was observed and, paralleled by a strong decrease in thein contrast to pMV158 and pMV(mob), maintenance of pMV158 derivatives. These*

pMVCm was only slightly unstable (more results confirm earlier observations thatthan 80% of the cells contained the plasmid accumulation of ss replicationafter 80 generations of growth; Fig. 5). intermediates of RCM plasmids areThese results indicate that the high associated with decreased plasmidinstability of pMV158 and pMV(mob) was maintenance in Gram-positive bacteria*

caused by either DNA sequences within the (Bron et al., 1987; Bron et al., 1988a; Bronmob gene, or by the N-terminal part of the et al., 1991; Gruss et al., 1987; Del Solar etMob protein which was not affected by the al., 1987; Gruss and Ehrlich, 1989). mutation in plasmid pMV(mob).*

DISCUSSION

A major conclusion from the presentwork is that several factors affect themaintenance of the rolling-circle, broad-host-range plasmid pMV158 in B.subtilis.First, the accumulation of high amounts ofssDNA was associated with severe plasmidlosses. Second, the palA-type SSO,although not an effective ssDNAconversion signal, had a positive effect onplasmid maintenance. Third, sequenceswithin the mob gene or part of the Mobprotein interfered with plasmidmaintenance. Another important conclusionis that the host-encoded RecA proteindrastically reduces ssDNA accumulation ofpMV158 derivatives lacking a functionalSSO and thereby improves plasmidmaintenance.

Effects of the pMV158 SSOs on ssDNAaccumulation and plasmid maintenance

The absence of a functional SSOfrom RCM plasmids results in theaccumulation of ssDNA (Del Solar et al.,1987; Boe et al., 1989; Gruss and Ehrlich,1989; Bron, 1990; Bron et al., 1991). InL.lactis, the accumulation of large amountsof ssDNA appeared to be a major factor

Alternative, RecA-dependent pathwa yfor ssDNA conversion

The present results also showed thatplasmids lacking a functional SSOaccumulated much more ssDNA inB.subtilis recA4 mutants than in wild-typerecA strains. This suggests that theB.subtilis RecA protein is involved in analternative, SSO-independent pathway ofssDNA conversion. Since the maintenanceof plasmids lacking a functional SSO wasmuch higher in the wild type recAbackground than in the recA4 mutant, thisalternative ssDNA conversion pathwayseems to be of considerable importance forthe maintenance of, at least, pMV158derivatives. Recently, we have obtainedevidence that, in addition to the host-encoded RecA protein, also the plasmidregion upstream of the double-strand origin(plus origin), which specifies an RNAtranscript complementary to the displacedssDNA strand, is important for thealternative ssDNA conversion pathway. Wehypothesize that the RecA proteinstimulates the annealing of the RNAtranscript to the displaced leading strandduring rolling-circle replication. Theannealed transcript can then serve as aprimer for lagging strand synthesis.Detailed analyses of this alternative ssDNAconversion route is described in chapter IV.The involvement of host-recombination

Chapter III

68

enzymes in plasmid maintenance has been effect of palA is masked by thereported before (Alonso et al., 1987; Viret destabilizing effects caused by the largeand Alonso, 1988). These authors showed amounts of ssDNA accumulated in thesethat pC194, lacking an SSO which is strains.functional in B.subtilis, is unstable in aB.subtilis recA4 background but not in awild type recA strain. These authors alsoshowed that pUB110, containing an SSOwhich is functional in B.subtilis, wasmaintained stably in both backgrounds(Alonso et al., 1987).

Posit ive effects of the palA SSO o nplasmid maintenance exclude the possibilities that the presence

Although pMV(U A ) and pMV(U of an intact Mob protein, or the joint- + -

A ) generated similar amounts of ssDNA, presence of palA and palU on the same-

the former was maintained markedly better plasmid, interfered with plasmidthan the latter in a wild-type recA maintenance. The results indicated thatbackground. Since the major difference B.subtilis cells containing pMV158between these plasmids is palA, this SSO, suffered from a severe growthwhich is not functional as an ssDNA disadvantage, resulting in a rapid outgrowthconversion signal in B.subtilis, is of pMV158-containing cells by theirapparently involved in the maintenance of plasmid-free progeny. Although we cannotpMV158. Del Solar et al. (1993) have entirely rule out the possiblity that the highrecently identified stretches of homology instability is caused by expression of partbetween part of palA and the partition (par) of the Mob protein [as pMV(mob ) was alsoregion of the E.coli plasmid pSC101. The highly unstable], the observation thathighest homology was found with the replacement of the pMV158 mob gene byregion containing the DNA gyrase binding other DNA sequences resulted in a drasticsite within this par site (Wahle and improvement of plasmid maintenance mostKornberg, 1988). Possibly, host-encoded probably indicates that sequences withinenzymes, such as DNA gyrase, bind to the the mob gene were responsible for the highpalA region, thus affecting plasmid instability of pMV158. So far, we have notsupercoiling and, as a consequence, plasmid delineated the sequences within mobmaintenance (Del Solar et al., 1993). That causing this instability in B.subtilis.altered levels of supercoiling can affect However, the effect cannot be caused byplasmid maintenance is known from several promoter activity of the mob gene, sincestudies with E.coli (Beaucage et al., 1991). this region is still intact in pMV(U A ) andA stabilizing effect of palA was also this plasmid is maintained stably. Althoughobserved in S.pneumoniae (Del Solar et al., the mechanism causing the instability of1993). Therefore, it is likely that palA, pMV158 in B.subtilis cannot be preciselybesides its ssDNA conversion activity in defined at present, it is clear that it is host-some Gram-positive bacteria, additionally dependent and that the type of mob gene iscan increase plasmid maintenance through, crucial, as pUB110 carrying a related mobmost likely, alterations in plasmid gene, is stably maintained in B.subtilis.configuration. However, since pMV(UA ) Taken together, the results described- +

is unstably maintained in B.subtilis recA4 in this chapter and those described instrains, we conceive that the stabilizing chapter II showed that, in addition to

Sequences within the pMV158 mob geneinterfere with plasmid maintenance

pMV158, containing both SSOs, wasmaintained stably in L.lactis (see chapterII) but not in B.subtilis. Since with pMV158no ssDNA was detectable, the observedinstability in B.subtilis must be aconsequence of other factors. We could

*

+ -

Chapter III

69

several other factors, the accumulation of growth rate of the cells, which consequentlylarge amounts of ssDNA strongly interfered results in rapid outgrowth of plasmid-with stable plasmid maintenance in both containing cells by plasmid-free cells. Thebacteria tested (L.lactis and B.subtilis). obtained results indicate that this is a likelySince it is not precisely known how ssDNA explanation for the observed instability ofaccumulation leads to plasmid loss, we can pMV(U A ) in L.lactis (see chapter II).only speculate about possible reasons forthis correlation.(a) It has been reported that the absence ofa functional SSO from certain plasmids inS.aureus (Gruss et al., 1987) andS.pneumoniae (Del Solar et al., 1987)results in a reduction in plasmid copynumber, which could be the cause of theobserved decrease in plamid maintenance inthose cases. However, such an explanationis unlikely to account for the observeddecrease in plasmid maintenance in ourstudies. First, no reduction in copy numberwas observed, and second, maintenance ofthe high-copy-number plasmid pMV(UA- -

)Em was not improved compared topMV(U A ).- -

(b) It is conceivable that ssDNAaccumulation results in, or increases thesize of, subpopulations of cells within theculture that have a higher than averageprobability of plasmid loss. This would bethe case if this subpopulation consists ofcells having a lower than average number ofsegregating plasmid units, for instance, dueto a higher spread in plasmid copy numbers.In this respect the recent work of Tolker-Nielsen and Boe (1994) is of interest. Theseauthours performed statistical analyses onthe formation of plasmid-free cells in E.colipopulations harbouring pBR322-derivedplasmids. A major conclusion was that thekinetics of plasmid loss did not fit into theconventional mathematical models due tothe presence of subpopulations of plasmid-containing cells which gave rise to progenythat produced plasmid-free cells with a highand unpredictable rate. So far, analyses of apossible effect of ssDNA accumulation onthe variation of plasmid copy numbers inindividual cells have not been published.(c) ssDNA accumulation might affect the

- -

MATERIALS AND METHODS

Bacterial strains, plasmids and media .Bacterial strains and plasmids used arelisted in Table 1. TY medium, used forculturing Escherichia coli and B.subtilis,contained Bacto tryptone (1%), Bacto yeastextract (0.5%) and NaCl (1%). TY platescontained in addition 2% agar.Tetracycline, kanamycin andchloramphenicol were added to finalconcentrations of 10 µg/ml.DNA techniques. All DNA manipulations,including the restriction of DNA, plasmidisolation, preparation of total lysates,purification of DNA fragments from gels,Southern hybridizations and thequantitation of ssDNA, were carried out asdescribed in chapter II.Transformation of B.subtilis and E.coli.Competent cells and protoplasts ofB.subtilis were prepared and transformed asdescribed (Bron, 1990). CaCl -treated2

E.coli cells were transformed as described(Sambrook et al., 1989).Copy number determinations an dassays of segregational plasmid stability.These procedures were as described inchapter II. The genome size of B.subtiliswas taken to be 4.12x10 bp (Amjad et al.,6

1991).Growth rates of plasmid-containing andplasmid-free cells. Two cultures oflogarithmically growing cells in selectivemedium, one with cells containing theplasmid and the other with plasmid-freecells, were mixed, and after centrifugationresuspended in non-selective medium. Theplasmid-free cells were distinguished from

Chapter III

70

Table 1. Plasmids and strains

Plasmid size a Relevant properties Copynumber b

Reference

pMV158 5.5 palA, palU, mob -gene, Tc R 15.1 ± 3.8 Burdett, 1980

pMV(U A )- + 4.4 palA, Tc R 15.7 ± 4.3 Lacks et al., 1986

pMV(U A )+ - 3.9 palU, Tc R 17,4 ± 4.9 This study

pMV(mob )* 5.5 palA, palU, inactive Mob,TcR

9.6 ± 3.6 This study

pMV(U A )- - 3.2 TcR 14.6 ± 5.3 This study

pMVCm 5.4 palA, palU, Î mob -gene, Tc ,R

CmR15.6 ± 4.2 This study

pMV(U A )Em- - 2.8 EmR >100 This study

pWUB10 5.4 palA , palU, inactive Mob,c

KmR50 e This study

pWUB20 5.4 palA , palU, inactive Mob,d

KmR50 e This study

pWUB11 5.4 palA , palU, inactive Mob,c

KmR5f This study

pWUB21 5.4 palA , palU, inactive Mob,d

KmR5f This study

Bacterial strain Genotype Reference

B.subtilis

8G5 trpC2 tyr1 met his ura nic ade rib , wild- typerecA

Bron and Venema, 1972

7G224 trpC2 tyr1 met ura nic ade rib recA4 de Vos and Venema,1981

E.coli

JM83 F- ara Î(lac -pro AB) rpsL (Str )r

(M80dÎ(lacZ)M15)Vieira and Messing,1982

a: Size in kb; b: Copy number per chromosome equivalent (average of 4 experiments); c: Same

plasmid-containing cells through a subsequent replica plating of the resultingchromosomally located kanamycin colonies on Km- and Tc-containing plates.resistance gene in the former strain. The During growth of the mixed cultures, theinitial ratio between plasmid-free and three possible cell types were distinguishedplasmid-containing cells was determined by by plating samples on non-selective plates,plating samples on non-selective media and replica plating the resulting coloniesdirectly after mixing the cultures and onto Km- and Tc-containing plates. To

Chapter III

71

maintain logarithmic growth of the cells themixed cultures were diluted in fresh,prewarmed, non-selective medium afterevery four hours.

ACKNOWLEDGEMENTS

We thank Henk Mulder for preparing theFigures. Funding for the project, of whichthis work is a part, was provided by STW(Stichting Technisch Wetenschappen, theNetherlands) and Gist-Brocades N.V.(Delft, The Netherlands). Part of this workwas supported by a CEC grant to S. Bron(Bridge program BIOT-CT910268).

REFERENCES

Alonso, J.C., Viret, J., and Tailor, R.H. (1987).Plasmid maintenance in Bacillus subtilisrecombination -deficient mutants. Mol .Gen. Genet. 208: 349-352.

Amjad, M., Castro, J.M., Sandoval, H., Wu ,J.-J ., Yang, M., Henner, D.J., an dPiggot, P.J. (1991). An Sfi I restrictio nmap of the Bacillus subtilis 168 genome.Gene 101: 15-21.

Beaucage, S.L., Miller, C.A., and Cohen, S.N .(1991). Gy rase-dependent stabilization ofpSC101 plasmid inheritance b ytranscri ptionally active promoters. EMBOJ. 10: 2583-2588.

Boe, L., Gerdes, K., and Molin, S. (1987). Effectsof genes exerting growth inhibition an dpl asmid stability on plasmi dma intenance. J. Bacteriol. 169:4646-4650.

Boe, L., Gros, M., Te Riele, H., Ehrlich, S.D. ,an d Gruss, A. (1989). Replicatio norigins of single-stranded-DNA plasmidpUB110. J. Bacteriol. 171: 3366-3372.

Bron, S., Bosma, P., Van Belkum, M.J., an dLu xen, E. (1987). Stability function i nthe Bacillus subtilis plasmid pTA1060 .Plasmid 18: 8-15.

Bron, S., Haima, P., Van Belkum, M.J., an dLuxen, E. (1988a). Segregationa lplasmid instability in Bacillus subtilis . InGen etics and Biotechnology of Bacilli .Ganesan, A.T. and Hoch, J.A. (eds.) San

Diego: Academic Press, Inc. pp. 305-309.Bron, S., Luxen, E., and Swart, P. (1988b) .

In stability of recombinant pUB11 0plasmids in Bacillus subtilis :pl asmid-encoded stability function an deffects of DNA inserts. Plasmid 19:231-241.

Bron, S. (1990). Plasmids. In Molecula rBiological Methods for Bacillus .Harwood, C.R. and Cutting, S.M. (eds. )Chichester, UK: John Wiley & Sons Ltd.pp. 75-174.

Bron, S., Meijer, W., Holsappel, S., and Haima,P. (1991). Plasmid instability an dmolecular cloning in Bacillus subtilis .Res. Microbiol. 142: 875-883.

Bron, S. and Venema, G. (1972). Ultraviole tina ctivation and excision repair i nBacillus subtilis . I. Construction an dcharacterization of a transformabl eeig htfold auxotrophic strain and tw oultraviolet-sensitive derivatives. Mutat .Res 15: 1-10.

Burdett, V. (1980). Identification of tetracyclineresistant R-plasmids in Streptococcu sagala ctiae (group B). Antimicrob. AgentsChemother. 18: 753-766.

de la Campa, A.G., Del Solar, G., and Espinosa,M. (1990). Initiation of replication o fplasmid pLS1. The initiator protein RepBacts on two distant DNA regions. J. Mol.Biol. 213: 247-262.

de Vos , W.M. and Venema, G. (1981). Fate o fplasmid DNA in transformation o fBacillus subtilis protoplasts. Mol. Gen .Genet. 182: 39-43.

del Solar, G., Kramer, G., Ballester, S., an dEspinosa, M. (1993). Replication of thepr omiscuous plasmid pLS1: a regio nen compassing the minus origin o fre plication is associated with stabl epla smid inheritance. Mol. Gen. Genet .241: 97-105.

del Solar, G.H., Puyet, A., and Espinosa, M .(19 87). Initiation signals for th econversion of single stranded to doubl estranded DNA forms in the streptococcalpl asmid pLS1. Nucleic Acids Res. 15:5561-5580.

del Solar, G.H., Pérez-Martin, J., and Espinosa,M. (1990). Plasmid pLS1-encoded RepAprotein regulates transcription fro mrepA B promoter by binding to a DN Asequence containing a 13-base pai rsymmetric element. J. Biol. Chem. 265:

Chapter III

72

12569-12575. Tol ker-Nielsen, T. and Boe, L. (1994). ADevi ne, K.M., Hogan, S.T., Higgins, D.G., an d sta tistical analysis of the formation o f

McConnell, D.J. (1989). Replication and pl asmid-free cells in populations o fsegregational stability of the Bacillu s Esc herichia coli . J. Bacteriol. 176:pl asmid pBAA1. J. Bacteriol. 171: 4306-4310. 1166-1172. van der Lelie, D., Bron, S., Venema, G., an d

Gruss, A. and Ehrlich, S.D. (1989). The family of Osk am, L. (1989). Similarity of minu shigly interrelated single-strande d origins of replication and flanking opendeoxyribonucleic acid plasmids . rea ding frames of plasmids pUB110 ,Microbiol. Rev. 53: 231-241. pTB913 and pMV158. Nucleic Acids Res.

Gruss, A .D., Ross, H.F., and Novick, R.P. (1987). 17: 7283-7294. Fu nctional analysis of a palindromi c va n der Lelie, D., Wösten, H.A.B., Bron, S. ,sequence required for normal replication Oskam, L., and Venema, G. (1990) .of seve ral staphylococcal plasmids. Proc. Conjugal mobilization of streptococca lNatl. Acad. Sci. USA 84: 2165-2169. pl asmid pMV158 between strains o f

Jannière, L., Gruss, A., and Ehrlich, S.D . Lactococcus lactis subsp. lactis . J.(1993). Plasmids. In Bacillus subtilis and Bacteriol. 172: 47-52. other Gram-positive bacteria ; Vieira, J. and Messing, J. (1982). The pU CBiochemistry, physiology and molecular plasmids, a M13mp7-derived system forgenetics. Sonenshein, A.L., Hoch, J.A. , insertion mutagenesis and sequencin gan d Losick, R. (eds.) Washington : with synthetic universal primers. Gen eAmerican Society for Microbiology, pp. 19: 259-268. 625-644. Viret, J. and Alonso, J.C. (1988). A DN A

Lacks, S.A., Lopez, P., Greenberg, B., an d sequence outside the pUB110 minima lEspinosa, M. (1986). Identification and replicon is required for norma lana lysis of genes for tetracyclin e rep lication in Bacillus subtilis . Nuclei cresistance and replication functions in the Acids Res. 16: 4389-4406. broad-host-range plasmid pLS1. J. Mol . Wah le, E. and Kornberg, A. (1988). Th eBiol. 192: 753-765. par tition locus of plasmid pSC101 is a

Leonhardt, H. (1990). Identification of a sp ecific binding site for DNA gyrase .lo w-copy-number mutation within th e EMBO J. 7: 1889-1895.pUB110 replicon and its effect o nplasmid stability in Bacillus subtilis .Gene 94: 121-124.

Meij er, W.J.J., van der Lelie, D., Venema, G. ,an d Bron, S. (1995). Effects of th egeneration of single-stranded DNA on themaintenance of plasmid pMV158 an dderivatives in Lactococcus lactis .Plasmid 33: 91-99.

No vick, R.P. (1989). Staphylococcal plasmid sand their replication. Annu. Rev .Microbiol. 43: 537-565.

Pérez-Martin, J. and Espinosa, M. (1991). TheRepA repressor can act as atranscriptional activator by inducin gDNA bends. EMBO J. 10: 1375-1382.

Priebe, S.D. and Lacks, S.A. (1989). Region o fth e streptococcal plasmid pMV15 8required for conjugative mobilization. J.Bacteriol. 171: 4778-4784.

Samb rook, J., Fritsch, E.F., and Maniatis, T .(1989). Molecular cloning: a laborator ymanual Cold Spring Harbor, New York :Cold Spring Harbor Laboratory Press.

Chapter III

73