Mobilization of “Nonmobilizable” Plasmids by the Aggregation-Mediated Conjugation System ofBacillus thuringiensis

PLASMID

36, 75–85 (1996) 0035

ARTICLE NO.

Mobilization of ‘‘Nonmobilizable’’ Plasmids by the Aggregation-Mediated Conjugation System of Bacillus thuringiensis LARS ANDRUP,1 OLE JØRGENSEN,* ANDREA WILCKS, LASSE SMIDT,

AND

GERT B. JENSEN

Department of Toxicology and Biology and *Department of Industrial Hygiene, National Institute of Occupational Health, Lersø Parkalle´ 105, DK-2100 Copenhagen, Denmark Received May 2, 1996; revised August 6, 1996 The aggregation-mediated conjugation system of Bacillus thuringiensis subsp. israelensis (Bti), encoded by the 200-kb plasmid pXO16, is highly potent in transferring itself and efficient in mobilizing other nonconjugative plasmids. In the present study we have analyzed the native Bacillus cereus plasmid pBC16. This plasmid has previously been shown to harbor a mob gene (ORFb) and a locus functioning as an oriT site in plasmid pLS20-mediated conjugation in Bacillus subtilis. However, in the conjugation system of Bti we found that a derivative of pBC16 deleted for both these loci was mobilizable, although at a reduced frequency. Another derivative of pBC16, containing a deletion spanning the first half of the coding region of the mob gene, was found to be nearly as mobilizable as the intact pBC16, suggesting its dispensability in the transfer process. Other plasmids based on the u-replicating origins, pAMb1, pLS20, ori43, ori44, and ori60, were also consistently mobilized in the conjugation system encoded by Bti plasmid pXO16. Analyzing the conjugation process by the use of scanning electron microscopy revealed the presence of connections between cells in the mating mixtures. These connections did not appear in monocultures of the donor strain or the recipient strain and may be conjugational junctions. q 1996 Academic Press, Inc.

Bacterial conjugation is a mechanism of genetic exchange that requires cell-to-cell contact and which is not susceptible to DNase present in the mating medium. Conjugation systems are encoded by large plasmids or by conjugative transposons (Clewell, 1993; Scott, 1993). Several conjugation systems in gram-positive bacteria have recently been discovered and characterized in some detail (for review see (Clewell, 1993)). The best studied is the pheromone-induced conjugation system of Enterococcus faecalis (Dunny et al., 1978), which along with the conjugation systems of Lactococcus lactis (van der Lelie et al., 1991), Lactobacillus plantarum (Reniero et al., 1992), and the aggregation-mediated conjugation system of the mosquito-toxic bacterium Bacillus thuringiensis subsp. israelensis (Bti)2

(Andrup et al., 1993; Jensen et al., 1995) are capable of sustaining DNA transfer in liquid media. These gram-positive conjugation systems differ from the gram-negative systems in that there are no pili involved in mating-pair formation. Other, yet still poorly understood, means of mating-pair formation and establishment of physical contact are employed here. The different conjugation systems in grampositive bacteria vary with respect to host range, requirement of solid surfaces, transfer frequencies, and ability to mobilize other DNA elements. The majority of conjugative plasmids reported in E. faecalis, Staphylococcus aureus, Staphylococcus epidermidis, and various bacilli can mobilize nonconjugative plasmids. However, differences in the ability of conjugative plasmids to mobilize nonconjugative plasmids were observed. Some nonconjugative plasmids could be mobilized by one conjugative plasmid and not by another. The conjugative plasmid pGO1 from S. aureus could not mobilize plasmids pT181 and

1 To whom correspondence should be addressed. Fax: /45 39 270107. E-mail: [email protected]. 2 Abbreviations used: Bti, Bacillus thuringiensis subsp. israelensis; oriT, origin of transfer; oriL, single-strand origin of replication.

75

0147-619X/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

AID

Plasmid 1257

/

6607$$$161

11-12-96 14:29:36

plasa

AP: Plasmid

76

ANDRUP ET AL.

pE194, whereas plasmids pE194 and pWBG3 (Projan and Archer, 1989), a plasmid similar to pT181, were mobilized by pWBG637 (Udo et al., 1992). Neither plasmid pC194 nor plasmid pE194 could be mobilized in the conjugation system of pLS20 in Bacillus subtilis (Selinger et al., 1990; Koehler and Thorne, 1987). Hence, it must be assumed that different mechanisms of DNA transfer are involved and that properties found in one system do not necessarily apply to another. In gram-negative bacteria, plasmid mobilization has been described to proceed through one of two mechanisms. Conduction is the process by which a self-transmissible plasmid causes transmission of a nonconjugative plasmid by physical association with it (Clark and Warren, 1979). Conduction usually results from recombinations between homologous sequences or is mediated by transposons on either the conjugative or the nonconjugative plasmid (Lambert et al., 1987). Donation, in contrast, is defined as the process whereby a nonconjugative plasmid is transferred via the apparatus of the conjugative plasmid, without physical association of the two plasmids (Clark and Warren, 1979). Donation requires some functions on the nonconjugative plasmid: a trans-acting mob gene (encoding a mobility protein) and a cis-acting origin of transfer (oriT) (for a review see (Lanka and Wilkins, 1995)). In S. aureus mobilization of the nonconjugative plasmid pC221 by means of the conjugative plasmid pGO1 is reported to be analogous to the mobilization by donation in gramnegative bacteria, though two trans-acting loci are required (Projan and Archer, 1989). Also, the conjugal mobilization of pBC16, a native Bacillus cereus plasmid (Bernhard et al., 1978), by pLS20 in B. subtilis is thought to take place by the process of donation (Koehler and Thorne, 1987). The Pre proteins of pT181 and pE194, which are closely related to the Mob proteins of pUB110, pTB913, and pMV158, act as site-specific recombinases at a palindromic site (RSA) during the formation of co-integrates (Gennaro et al., 1987; Oskam et al., 1991). The RSA site of pUB110 has

AID

Plasmid 1257

/

6607$$$162

been shown to be involved in conjugative mobilization and it is suggested that the potential palindromic structure of the RSA site has a similar function as the gram-negative oriT sites (Oskam et al., 1991; Selinger et al., 1990; Priebe and Lacks, 1989; Willetts and Wilkins, 1984) and that mobilization may proceed via site-specific recombination (Reimmann and Haas, 1993). We have studied the conjugation-like system of Bti first discovered by Gonza´lez and Carlton (1984). Recently, we found that mobilization of small plasmids between strains of Bti is accompanied by non-pheromone-induced and protease-sensitive coaggregation between donor and recipient cells (Andrup et al., 1993). Further, the genetic basis of this aggregation system (the Agr/ phenotype) has been shown to reside on pXO16, a 200-kb conjugative plasmid (Jensen et al., 1995). By transposon tagging of plasmid pXO16, it was demonstrated to virtually transfer to every recipient cell in short-term broth matings and it was shown that this aggregation-mediated conjugation system was fully functional in several B. thuringiensis subspecies and in B. cereus (Jensen et al., 1996). In the present study our goal was to analyze the mobilization process of, primarily, the native B. cereus plasmid pBC16 (Bernhard et al., 1978) on which a mob gene and a site (RSA) suggested to function as origin of transfer (oriT) have been identified (Selinger et al., 1990). Results from previous work have suggested that the mobilization mechanisms of the aggregation-mediated conjugation system of Bti may be of a different nature than those other systems so far examined. We have previously reported the mobilization of Bti plasmid pTX14-3 and derivatives containing only the replicon of the plasmid (Andrup et al., 1993, 1995), suggesting that the replicon of pTX14-3 is sufficient to sustain mobilization in this conjugation system. MATERIALS AND METHODS Strains, Media, and Plasmids Bti strains used in this study are listed in Table 1. The derivatives of pBC16 were con-

11-12-96 14:29:36

plasa

AP: Plasmid

77

MOBILIZATION IN Bacillus thuringiensis TABLE 1 BACTERIAL STRAINS Strain or plasmid Strains AND508 GBJ001 AND801 Plasmids pXO16 pBC16 pBC16DEcoRI pBC16DNdeI pBC16invEcoRI pBC16DBsiHKAI pC194 pE194 pTLS6D4 pWKM6 pEG599 pEG588-8 pEG588-14a pHV1432

Plasmid 1257

PLASMIDS

Relevant characteristics

Reference

B. thuringiensis subsp. israelensis cured of the small plasmids, Agr/ Bti strain cured of all plasmids, Str r, Agr0 Bti strain GBJ001 containing plasmid pXO16, Str r, Agr/

Andrup et al. (1993) Jensen et al. (1995) Jensen et al. (1995)

Natural conjugative Bti plasmid; 200 kb Natural B. cereus plasmid, Tet r pBC16 with a 1574-bp EcoRI deletion pBC16 with a 498-bp NdeI deletion pBC16 with the 1574-bp EcoRI fragment inverted pBC16 with a 129-bp BsiHKAI deletion Natural Staphylococcus aureus plasmid, Camr Natural S. aureus plasmid, Ermr pSC101-based vector containing 2.9 kb from pLS20, Tet r pUC7-based vector containing minimal replicon from pLS20, Kanr pTZ18u-based vector containing ori43 from B. thuringiensis HD263, Camr pTZ18u-based vector containing ori44 from B. thuringiensis HD263, Camr pTZ18u-based vector containing ori60 from B. thuringiensis HD263, Camr Derivative of pHV1431 containing the replicon of pAMb1, Camr

structed by digesting pBC16 with restriction enzymes as indicated in Table 1, ligating the mixture, and electroporating it into Bti strain AND508. The correct plasmid derivatives were identified among the electrotransformants by the use of restriction enzyme analysis using ClaI, PvuII, BamHI, NdeI, and EcoRI. Plasmids pE194 and pC194 were obtained from Per Lina˚ Jørgensen, Novo-Nordisk A/S, Copenhagen, plasmids pTLS6D4 and pWKM6 were obtained from Wilfried J. J. Meijer, Groningen Biomolecular Sciences and Biotechnology Institute, Haren, and plasmids pEG599, pEG588-8, and pEG588-14a were provided by James A. Baum, Ecogen Inc., Pennsylvania. All cultures were grown in LB media (Sambrook et al., 1989) supplemented with antibiotics (Sigma), when appropriate, as follows: 10 mg/ml of kanamycin, 6 or 10 mg/ml of chloramphenicol, 6 mg/ml of tetracycline (when selecting transconjugants)

AID

AND

/

6607$$$162

Jensen et al. (1995) Bernhard et al. (1978) This study This study This study This study Ehrlich (1977) Iordanescu (1976) Meijer et al. (1995) Meijer et al. (1995) Baum et al. (1990) Baum et al. (1990) Baum et al. (1990) Simon and Chopin (1988)

and 10 mg/ml of tetracycline (in other cases), 10 mg/ml of erythromycin, and 100 mg/ml of streptomycin. Restriction Enzymes and Molecular Cloning Restriction endonucleases were purchased from GIBCO-BRL, Life Technologies and used as recommended by the supplier. Molecular cloning techniques were performed as described by Sambrook et al. (1989). Electroporation of Bti was performed as follows: 400 ml exponential culture (OD600 É 0.2) was harvested, washed once with 200 ml ice-cold water and twice with ice-cold 10% glycerol, and resuspended in 1 ml of ice-cold 10% glycerol. For each electroporation, 150 ml of cells was mixed with 1 to 3 ml of DNA (miniprep or ligation mixture) and subjected to a single pulse of 5.5 kV/cm (200 V, 21 mF) in a 2mm cuvette (Bio-Rad) using the geneZapper

11-12-96 14:29:36

plasa

AP: Plasmid

78

ANDRUP ET AL.

electroporation system (International Biotechnologies, Inc.). The cells were washed out of the cuvette with 1 ml prewarmed LB medium and incubated at 307C with shaking for up to 24 h before being plated on selective media. Plasmid DNA Extraction and Analysis Plasmid DNA from Bti, predominantly containing plasmids smaller than 15 kb, was extracted by a modification of the alkaline lysis method described previously (Andrup et al., 1993). DNA was analyzed by horizontal gel electrophoresis (6–10 V/cm) in 0.8 or 1% agarose (SeaKem GTG) with TBE buffer (Sambrook et al., 1989). The minigels were stained after electrophoresis in 1 mg/ml of ethidium bromide for 20 min and destained in water. Mating in Broth Overnight cultures of the donor and recipient strains, grown separately at 307C in LB medium containing the appropriate antibiotics, were diluted 1:100 into fresh prewarmed (307C) LB medium without antibiotics. Equal amounts (250 ml per OD600 unit or as indicated) of recipient and donor cells in late logarithmic growth were combined in 7 ml prewarmed LB medium and incubated at 307C with moderate shaking (180 rpm). After 3 h, dilutions were plated on appropriate selective media for determining the number of donors, recipients, and transconjugants. Controls of donors and recipients grown separately were also included. Plasmid Stability In order to determine the segregational stability of pBC16 and its derivatives, cultures of strain AND508 containing the plasmid in question were kept exponentially growing at 307C by repetitive dilutions in prewarmed medium. After 30 h, serial dilutions were spread on agar plates without antibiotics. After incubation overnight at 307C, individual colonies were toothpicked to agar plates containing tet-

AID

Plasmid 1257

/

6607$$$162

racycline and the frequency of plasmid loss was calculated. Scanning Electron Microscopy Preparations for scanning electron microscopy were done by using the method of Jiang and Chai (1996) with some alterations. Monocultures of strain AND801, strain GBJ001, and mating mixtures comprising strains AND801 (Agr/) and GBJ001 (Agr0) were harvested by centrifugation. The cell pellets were washed twice with 11 phosphate-buffered saline (Sambrook et al., 1989). The cells were fixed for 1 h at room temperature with 2% (vol/vol) glutaraldehyde solution buffered at pH 7.4 with 0.1 M cacodylate. The samples were passed through 13-mm-diameter, 0.2mm-pore-size Nuclepore polycarbonate filters (Costar, Cambridge, MA) and then postfixed in 1% (wt/vol) osmium tetroxide using the same buffer. After postfixation the cells were dried with a graded ethanol series [75, 96, and 100% (vol/vol)]. The samples were coated with Au in a vacuum evaporator coating apparatus, Model E5000 (Kjellbergs Successors AB, Sweden). The coated samples were examined in a scanning electron microscope (JEM EX1200II/STEM, Jeol, Japan) at 10 kV. RESULTS AND DISCUSSION

Broth matings between strains of Bti consistently showed, when using a transposontagged derivative of plasmid pXO16, a transfer frequency of 100% (all recipients had acquired pXO16::Tn5401) (Jensen et al., 1996). However, when mobilizing nonconjugative plasmids the transfer frequencies obtained varied considerably between different mating experiments mobilizing the same plasmid. Therefore, a study was undertaken to determine some of the main parameters affecting the transfer frequencies of plasmid pBC16. Plasmid mobilization experiments were conducted using strain AND508 containing plasmid pBC16 as donor and the streptomycin-resistant strain GBJ001 as recipient. It was found that the ratio of donors to recipients was important for the transfer frequency deter-

11-12-96 14:29:36

plasa

AP: Plasmid

79

MOBILIZATION IN Bacillus thuringiensis

FIG. 1. Correlation between the frequency of transfer (transconjugants/recipient) of plasmid pBC16 and the ratio of donors to recipients in the mating mixtures. The ratio of donors to recipients was determined at the end of each mating as the number of tetracycline-resistant cells divided by the number of streptomycinresistant cells.

mined. The donor and recipient cells were combined at various ratios and the matings were performed as described under Materials and Methods. The appearance of visible aggregates in the mating mixtures about 15–20 min after the combination of donor and recipient cells was recognized. The ‘‘quality,’’ judged from the size of the aggregates and the number of cells not involved in aggregation (background turbidity), however, was heterogeneous. As seen in Fig. 1 there was a correlation between donor to recipient ratio and the transfer frequency measured. The rationale for this phenomenon may be that in a given mating event, pBC16 is only transferred to a limited extent, whereas the conjugative plasmid pXO16 is transferred at a frequency of 100%. Therefore, when the donor:recipient ratio is low, an increasing fraction of newly formed donors is only able to transfer the conjugative pXO16; these are the recipients that have acquired only plasmid pXO16. Because of entry exclusion and aggregation incompatibility, the recipients will be saturated with pXO16, resulting in a decreased transfer frequency of pBC16. On the other hand, when the donor:recipient ratio is high there is a higher probability that recipients form mating pairs with the original donors actually containing pBC16, resulting in a higher transfer frequency.

AID

Plasmid 1257

/

6607$$$163

Plasmid pBC16, conferring tetracycline resistance, and pUB110, a S. aureus plasmid specifying kanamycin resistance, are incompatible and highly homologous except for the region occupied by their resistance determinants (Polak and Novick, 1982). Recently, the access to the complete DNA sequence of pBC16 (Tang and Wilson, 1995) has made an exact analysis of the homology possible. It appears that only one base pair differed in the two plasmids outside the region occupied by their resistance genes (an A is located at position 4155 of the pUB110 sequence, whereas a G is reported at position 4237 of pBC16). The ORFb (mob gene) of both pUB110 and pBC16 has been shown to be essential for mobilization by plasmid pLS20 in B. subtilis and a region upstream of this gene, presumably the RSA site, has also been demonstrated to be essential for mobilization in the abovementioned conjugation system (Selinger et al., 1990). We constructed three derivatives of pBC16. The first plasmid, pBC16DNdeI, contains a deletion of a 498 bp NdeI fragment located in the first part of the mob gene. The second construct has inverted the 1574-bp EcoRI fragment, resulting in a truncated mob gene and an inversion of the RSA site and the single-strand origin of replication (oriL). The third derivative of plasmid pBC16 has deleted

11-12-96 14:29:36

plasa

AP: Plasmid

80

ANDRUP ET AL.

FIG. 2. Restriction and genetic map of pBC16. An EcoRI fragment (indicated as a dashed line) is deleted in plasmid pBC16DEcoRI and inverted in pBC16invEcoRI. A NdeI fragment (indicated as a dotted line) is deleted in plasmid pBC16DNdeI. Numbers in parenthesis refer to the sequence position as determined by Tang and Wilson (1995).

the 1574-bp EcoRI fragment containing the main part of the mob gene, the RSA site, and the oriL (Fig. 2). Plasmid pBC16 and its derivatives were inserted into strain AND508 by electroporation and the resulting strains were used as donors. No differences in copy numbers, judged from gel electrophoresis of minipreps, were observed and the strains containing pBC16 or the derivatives grew at comparable growth rates (data not shown). Figure 3 shows the transfer frequencies of pBC16 and its derivatives. A streptomycin-resistant Bti strain was used as recipient. Each mating was conducted at least three times. Apparently, the knockout of the mob gene did not significantly reduce the transfer frequency, as plasmid pBC16DNdeI was mobilized nearly as efficiently as pBC16. However, deleting or inverting the EcoRI fragment resulted in a 10fold reduction in the transfer frequency. In the plasmid design used, it was not possible to differentiate the activity of the RSA (positions 1145 to 1168, Priebe and Lacks, 1989, Fig. 2) site and oriL (positions 1246 to 1522, Boe

AID

Plasmid 1257

/

6607$$$163

et al., 1989, Fig. 2) directly. However, it is generally accepted that an origin of transfer (oriT) is fully proficient in either orientation, at least in gram-negative conjugation systems. If the RSA site, as suggested by Sellinger et al. in pLS20-mediated conjugation (1990), should function as oriT then inverting the EcoRI fragment should not impair its activity. Nonetheless, the intact mob gene is not essential for a high mobilization frequency, but the orientation of the sequences upstream this gene seems to be crucial. Hence, if the RSA site is providing the mobilization activity it is not functioning in the same way as the oriT loci analyzed in gram-negative systems. On the other hand, single-strand origins of replication must be located on the same strand as the double-strand origin (Seery et al., 1993; Bron, 1990); therefore, the mobilization activity, dependent on the orientation of the EcoRI fragment, may be attributed to the oriL locus. The higher frequency of mobilization exhibited by plasmids pBC16 and pBC16DNdeI compared to pBC16DEcoRI and pBC16in-

11-12-96 14:29:36

plasa

AP: Plasmid

81

AID

Plasmid 1257

/

6607$$$163

11-12-96 14:29:36

plasa

The transfer frequencies (transconjugants pr recipient) are averages of six mating experiments ({SD). Donor strain AND508 containing the plasmid(s) indicated. c Transconjugants selected on streptomycin and tetracycline. d Transconjugants selected on streptomycin and chloramphenicol. e Transconjugants selected on streptomycin and erythromycin. f Transconjugants selected on streptomycin, tetracycline, and chloramphenicol. g Transconjugants selected on streptomycin, tetracycline, and erythromycin. b

a

1.8 1 1004 { 1.1 1 1004 4.3 1 1005 { 2.2 1 1005

2.5 1 1004 { 1.6 1 1004 4.2 1 1003 { 2.0 1 1003

pBC16 pC194 pE194 pBC16 / pC194 pBC16 / pE194

3.0 1 1003 { 9.3 1 1004 9.9 1 1004 { 6.1 1 1004

6.4 1 1005 { 1.9 1 1005

pE194e pC194d pBC16c Plasmid(s) in donor b

TABLE 2

PLASMIDS CORESIDENT OF

MOBILIZATION

vEcoRI cannot solely be ascribed to the higher stability of the plasmids. Plasmid pBC16DNdeI is as stable as plasmid pBC16 (ú99% stability after 60 generations), whereas the constructs containing the EcoRI deletion or inversion are less stable (90 and 75% stability after 60 generations, respectively). Presumably, this effect is caused by the deleted or inverted single-strand origin of replication located on the EcoRI fragment. If single-strand transfer is assumed, the 10-fold reduction of the transfer frequencies observed for the plasmid derivatives with impaired single-strand origin might be ascribed to difficulties in establishing in the recipient cell. To substantiate the indication that the function of the single-strand origin of replication is important for the number of transconjugants detected, we constructed a derivative of pBC16 (pBC16DBsiHKAI) with a small (129 bp, see Fig. 2) deletion containing the primary site of lagging strand replication (nt 1299 { 1, Dempsey et al., 1995). Mobilizing this derivative verified that an intact single-strand origin is necessary for high frequency of transfer, i.e., pBC16DBsiHKAI was transferred at a frequency of less than half that of pBC16 (data not shown).

IN THE

FIG. 3. Mobilization frequencies, measured as transconjugants per donor, for pBC16, its derivatives, pC194, and pE194. Each experiment was repeated at least four times. The standard deviations are indicated on top of each bar.

Transfer frequency of plasmid(s)a

DONOR STRAIN

pBC16 / pC194 f

7.6 1 1007 { 9.4 1 1007

pBC16 / pE194g

3.5 1 1006 { 6.5 1 1007

MOBILIZATION IN Bacillus thuringiensis

AP: Plasmid

82

ANDRUP ET AL.

We have analyzed several transconjugants from each mating for their contents of small plasmids. They had all acquired pBC16 or the derivative in question and, inferred from their restriction pattern, had not recombined or deleted. This suggests that mobilization of these plasmids by the aggregation-mediated conjugation system of plasmid pXO16 takes place through the process of donation. Even though the inversion or deletion of the EcoRI fragment of pBC16 resulted in a reduced transferability, the various derivatives were still consistently mobilized. We have previously reported the mobilization of a native Bti plasmid pTX14-3 (Andrup et al., 1993, 1995). Mobilizing derivatives of plasmid pTX14-3 demonstrated that the region containing the rep gene is sufficient to sustain DNA transfer when pXO16 is coresident in the donor strain. To establish the extent of the mobilizing abilities of plasmid pXO16, we performed mating experiments with donor strain AND508 containing either plasmid pE194 or plasmid pC194, both originating from S. aureus (Iordanescu, 1976; Iordanescu and Surdeanu, 1980). The reasons for choosing these plasmids are: (i) they are common cloning vectors, (ii) they have been reported not to be mobilizable in the conjugation system of pLS20 of B. subtilis (Selinger et al., 1990; Koehler and Thorne, 1987), and (iii) pE194 contains the pre gene with extensive homology to the mob gene of pUB110 and pBC16 and an RSA site which is nearly identical (one mismatch) to the RSA site of pUB110 and pBC16 (Gennaro et al., 1987). As shown in Fig. 3, both pE194 and pC194 were consistently mobilized. The frequencies were comparable and of approximately the same order as transfer frequencies of the pBC16 derivatives containing a deletion or an inversion of the EcoRI fragment. It has been demonstrated that the single-strand origins of plasmids pC194 and pE194, as opposed to the single-

strand origins of plasmids pUB110 and pBC16, do not function in B. subtilis (Gruss et al., 1987; Boe et al., 1989). This fact may explain their lower frequencies of transfer in Bti than pBC16 containing a functioning single-strand origin. In order to see whether the Mob protein of pBC16 would have any influence on the transfer of pC194 or pE194, we constructed donor strains containing both pBC16 and either pC194 or pE194 coresident. The results of the five mating experiments, repeated six times, are shown in Table 2. It should be mentioned that the chloramphenicol concentration used for the selection of plasmid pC194 was 10 mg/ ml, whereas 6 mg/ml was used in the experiments shown in Fig. 3. We have previously described the correlation between chloramphenicol concentration and the number of transconjugants obtained following a mating (Andrup et al., 1995); this is the reason for the reduced transfer frequency measured in the matings shown in Table 2 compared to those in Fig. 3. It seems obvious that there was no enhanced mobilizability exerted by the coresident plasmid pBC16. On the contrary, the transfer of both plasmids seemed to be reduced. The effect is most pronounced for the combination of pE194 and pBC16; the reason for this is not known. However, the Mob protein of pBC16 and the Pre protein of pE194 show a high degree of homology and may interfere negatively. Alternatively, this effect may be a result of altered stability of the plasmids. Interestingly, a relatively high rate of cotransfer of both plasmids was observed in both matings. The result of these experiments showed that the aggregation-mediated aggregation system of Bti is able to mobilize any plasmids coresident with the conjugative aggregation plasmid pXO16. To further substantiate this observation, we performed matings to ascertain whether it would be possible to mobilize plas-

FIG. 4. Scanning electron micrographs of a mating mixture comprising donor cells (strain AND801) and recipient cells (strain GBJ001). These connections were abundantly found in mating mixtures but not in monocultures of either the donor strain or the recipient strain. Bar, 1 mm.

AID

Plasmid 1257

/

6607$$$163

11-12-96 14:29:36

plasa

AP: Plasmid

83

MOBILIZATION IN Bacillus thuringiensis

AID

Plasmid 1257

/

6607$$1257

11-12-96 14:29:36

plasa

AP: Plasmid

84

ANDRUP ET AL.

mids based on non-rolling-circle replicating replicons. Conjugative replication of the Fplasmid is known to take place by a process similar to rolling-circle replication used by most small plasmids of gram-positive bacteria (Willetts and Wilkins, 1984), and this replication mechanism has previously been suggested to be involved in the horizontal DNA transfer of small plasmids (Dempsey et al., 1995; Gennaro et al., 1987). Two plasmids, pWKM6 and pTLS6D4, containing the replicon of the conjugative B. subtilis plasmid pLS20 (Koehler and Thorne, 1987; Meijer et al., 1995), three plasmids based on large B. thuringiensis plasmids, pEG588-8, pEG58814a, pEG599 (Baum et al., 1990), and a plasmid, pHV1432 (Simon and Chopin, 1988), containing the replicon of the broad-hostrange plasmid pAMb1 were analyzed. They were all mobilizable. Plasmids pWKM6, pTLS6D4, pEG588-14a, and pHV1432 were mobilized at frequencies between 1 and 7 1 1005, whereas plasmids pEG588-8 and pEG599 were transferred at frequencies at about 1 1 1004. A scanning electron microscopy study was undertaken. Mating mixtures and monocultures of donor and recipient cells were examined in a scanning electron microscope. Structures like the ones shown in Fig. 4 were easily found in mating mixtures, but not in the monocultures. Whether these connections are in fact conjugational junctions or structures for bringing the cells into contact with each other we do not know. Often we found that one cell was ‘‘connected’’ to several other cells at a time and we infer that these connections are the basis of the macroscopic aggregation. The fact that all plasmids tested so far could be mobilized in the pXO16-encoded conjugation system of B. thuringiensis suggests that this is an exceptional, and so far unique, system. Both rolling-circle replicating plasmids and plasmids based on u-replicating origins could be mobilized apparently without significant differences. Some plasmids, such as pBC16, the derivative pBC16DNdeI, pEG599, and pEG588-8, were transferred at higher frequencies in these short-term mat-

AID

Plasmid 1257

/

6607$$$163

ings. Whether this can be attributed to stability, replication mechanisms, conformation, or size has yet to be clarified. It may be speculated that the mechanism of mobilization, mediated by the conjugative aggregation plasmid pXO16, involves cytoplasmatic fusions between aggregating cells and that the transfer of nonconjugative plasmids is the result of random migration through conjugation channels rather than a more ‘‘guided’’ mobilization. ACKNOWLEDGMENTS We are grateful to Claus A. Jarløv for his linguistic assistance. We thank Esben Kjær Sørensen and Annette Rasmussen for technical assistance and Claus Thor Nielsen for managing the references. Wilfried J. J. Meijer, James Baum, and Per Lina˚ Jørgensen are thanked for their donation of strains and plasmids. L.A. was supported by a grant from the Nordic Council of Ministers and G.B.J. was supported by the Danish Working Environment Fund.

REFERENCES ANDRUP, L., DAMGAARD, J., AND WASSERMANN, K. (1993). Mobilization of small plasmids in Bacillus thuringiensis subsp. israelensis is accompanied by specific aggregation. J. Bacteriol. 175, 6530–6536. ANDRUP, L., BENDIXEN, H. H., AND JENSEN, G. B. (1995). Mobilization of Bacillus thuringiensis plasmid pTX143. Plasmid 33, 159–167. BAUM, J. A., COYLE, D. M., GILBERT, M. P., JANY, C. S., AND GAWRON-BURKE, C. (1990). Novel cloning vectors for Bacillus thuringiensis. Appl. Environ. Microbiol. 56, 3420–3428. BERNHARD, K., SCHREMPF, H., AND GOEBEL, W. (1978). Bacteriocin and antibiotic resistance plasmids in Bacillus cereus and Bacillus subtilis. J. Bacteriol. 133, 897– 903. BOE, L., GROS, M. F., TE RIELE, H., EHRLICH, S. D., AND GRUSS, A. (1989). Replication origins of singlestranded-DNA plasmid pUB110. J. Bacteriol. 171, 3366–3372. BRON, S. (1990). Plasmids. In ‘‘Molecular Biology Methods for Bacillus’’ (C. R. Harwood and S. M. Cutting, Eds.), pp. 75–138. Wiley, New York. CLARK, A. J., AND WARREN, J. (1979). Conjugal transmission of plasmids. Annu. Rev. Genet. 13, 99–125. CLEWELL, D. B. (1993). ‘‘Bacterial Conjugation.’’ Plenum, New York. DEMPSEY, L. A., ZHAO, A. C., AND KHAN, S. A. (1995). Localization of the start sites of lagging-strand replication of rolling-circle plasmids from gram-positive bacteria. Mol. Microbiol. 15, 679–687. DUNNY, G. M., BROWN, B. L., AND CLEWELL, D. B.

11-12-96 14:29:36

plasa

AP: Plasmid

85

MOBILIZATION IN Bacillus thuringiensis (1978). Induced cell aggregation and mating in Streptococcus faecalis: evidence for a bacterial sex pheromone. Proc. Natl. Acad. Sci. USA 75, 3479–3483. EHRLICH, S. D. (1977). Replication and expression of plasmids from Staphylococcus aureus in Bacillus subtilis. Proc. Natl. Acad. Sci. USA 74, 1433–1436. GENNARO, M. L., KORNBLUM, J., AND NOVICK, R. P. (1987). A site-specific recombination function in Staphylococcus aureus plasmids. J. Biotechnol. 169, 2601– 2610. GONZALEZ, J. M., JR., AND CARLTON, B. C. (1984). A large transmissible plasmid is required for crystal toxin production in Bacillus thuringiensis variety israelensis. Plasmid 11, 28–38. GRUSS, A., ROSS, H., AND NOVICK, R. (1987). Functional analysis of a palindromic sequence required for normal replication of several staphylococcal plasmids. Proc. Natl. Acad. Sci. USA 84, 2165–2169. IORDANESCU, S. (1976). Three distinct plasmids originating in the same Staphylococcus aureus strain. Arch. Roum. Pathol. Exp. Microbiol. 35, 111–118. IORDANESCU, S., AND SURDEANU, M. (1980). New incompatibility groups for Staphylococcus aureus plasmids. Plasmid 4, 256–260. JENSEN, G. B., WILCKS, A., PETERSEN, S. S., DAMGAARD, J., BAUM, J. A., AND ANDRUP, L. (1995). The genetic basis of the aggregation system in Bacillus thuringiensis subsp. israelensis is located on the large conjugative plasmid pXO16. J. Bacteriol. 177, 2914–2917. JENSEN, G. B., ANDRUP, L., WILCKS, A., SMIDT, L., AND POULSEN, O. M. (1996). The aggregation-mediated conjugation system of Bacillus thuringiensis subsp. israelensis: Host range and kinetics. Curr. Microbiol. 33, 228–236. JIANG, X., AND CHAI, T. (1996). Survival of Vibrio parahaemolyticus at low temperatures under starvation conditions and subsequent resuscitation of viable, nonculturable cells. Appl. Environ. Microbiol. 62, 1300–1305. KOEHLER, T. M., AND THORNE, C. B. (1987). Bacillus subtilis (natto) plasmid pLS20 mediates interspecies plasmid transfer. J. Bacteriol. 169, 5271–5278. LAMBERT, C. M., HYDE, H., AND STRIKE, P. (1987). Conjugal mobility of the multicopy plasmids NTP1 and NTP16. Plasmid 18, 99–110. LANKA, E., AND WILKINS, B. M. (1995). DNA processing reactions in bacterial conjugation. Annu. Rev. Biochem. 64, 141–169. MEIJER, W. J. J., DE BOER, A. J., VAN TONGEREN, S., VENEMA, G., AND BRON, S. (1995). Characterization of the replication region of the Bacillus subtilis plasmid pLS20: a novel type of replicon. Nucleic Acids Res. 23, 3214–3223. OSKAM, L., HILLENGA, D. J., VENEMA, G., AND BRON, S. (1991). The large Bacillus plasmid pTB19 contains two

AID

Plasmid 1257

/

6607$$$164

integrated rolling-circle plasmids carrying mobilization functions. Plasmid 26, 30–39. POLAK, J., AND NOVICK, R. P. (1982). Closely related plasmids from Staphylococcus aureus and soil bacilli. Plasmid 7, 152–162. PRIEBE, S. D., AND LACKS, S. A. (1989). Region of the streptococcal plasmid pMV158 required for conjugative mobilization. J. Bacteriol. 171, 4778–4784. PROJAN, S. J., AND ARCHER, G. L. (1989). Mobilization of the relaxable Staphylococcus aureus plasmid pC221 by the conjugative plasmid pGO1 involves three pC221 loci. J. Bacteriol. 171, 1841–1845. REIMMANN, C., AND HAAS, D. (1993). Mobilization of chromosomes and nonconjugative plasmids by cointegrative mechanisms. In ‘‘Bacterial Conjugation’’ (D. B. Clewell, Ed.), pp. 137–188. Plenum, New York. RENIERO, R., COCCONCELLI, P., BOTTAZZI, V., AND MORELLI, L. (1992). High frequency of conjugation in Lactobacillus mediated by an aggregation-promoting factor. J. Gen. Microbiol. 138, 763–768. SAMBROOK, J., FRITSCH, E. F., AND MANIATIS, T. (1989). ‘‘Molecular Cloning: A Laboratory Manual,’’ 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. SCOTT, J. R. (1993). Conjugative transposons. In ‘‘Bacillus subtilis and Other Gram-Positive Bacteria’’ (A. L. Sonenshein, J. A. Hoch, and R. Losick, Eds.), pp. 597– 614. Am. Soc. for Microbiol., Washington, DC. SEERY, L. T., NOLAN, N. C., SHARP, P. M., AND DEVINE, K. M. (1993). Comparative analysis of the pC194 group of rolling circle plasmids. Plasmid 30, 185–196. SELINGER, L. B., MCGREGOR, N. F., KHACHATOURIANS, G. G., AND HYNES, M. F. (1990). Mobilization of closely related plasmids pUB110 and pBC16 by Bacillus plasmid pXO503 requires trans-acting open reading frame b. J. Bacteriol. 172, 3290–3297. SIMON, D., AND CHOPIN, A. (1988). Construction of a vector plasmid family and its use for molecular cloning in Streptococcus lactis. Biochimie 70, 559–566. TANG, M., AND WILSON, C. R. (1995). The complete sequence of plasmid pBC16 and its relationship to plasmid pUB110. Proceedings of the 8th International Conference on Bacilli 74. [Abstract] UDO, E. E., LOVE, H., AND GRUBB, W. B. (1992). Intraand inter-species mobilisation of non-conjugative plasmids in staphylococci. J. Med. Microbiol. 37, 180– 186. VAN DER LELIE, D., CHAVARRI, F., VENEMA, G., AND GASSON, M. J. (1991). Identification of a new genetic determinant for cell aggregation associated with lactose plasmid transfer in Lactococcus lactis. Appl. Environ. Microbiol. 57, 201–206. WILLETTS, N., AND WILKINS, B. (1984). Processing of plasmid DNA during bacterial conjugation. Microbiol. Rev. 48, 24–41. Communicated by S. Khan

11-12-96 14:29:36

plasa

AP: Plasmid