SGM Journals
Research Article

The tra locus of streptomycete plasmid pIJ101 mediates efficient transfer of a circular but not a linear version of the same replicon

Jing Wang, Gregg S. Pettis

  • Department of Biological Sciences, Louisiana State University, Baton Rouge, LA 70803, USA
  • Correspondence
    Gregg S. Pettis
    gpettis{at}lsu.edu

    • Microbiology 2010; 156(9):2723–2733 · https://doi.org/10.1099/mic.0.036467-0

    View at publisher PubMed

    Abstract

    Conjugal transfer of circular plasmids in Streptomyces involves a unique mechanism employing few plasmid-encoded loci and the transfer of double-stranded DNA by an as yet uncharacterized intercellular route. Efficient transfer of the circular streptomycete plasmid pIJ101 requires only two plasmid loci: the pIJ101 tra gene, and as a cis-acting function known as clt. Here, we compared the ability of the pIJ101 transfer apparatus to promote conjugal transfer of circular versus linear versions of the same replicon. While the pIJ101 tra locus readily transferred the circular form of the replicon, the linear version was transferred orders of magnitude less efficiently and all plasmids isolated from the transconjugants were circular, regardless of their original configuration in the donor. Additionally, relatively rare circularization of linear plasmids was detectable in the donor cells, which is consistent with the notion that this event was a prerequisite for transfer by TraB(pIJ101). Linear versions of this same replicon did transfer efficiently, in that configuration, from strains containing the conjugative linear plasmid SLP2. Our data indicate that functions necessary and sufficient for transfer of circular DNA were insufficient for transfer of a related linear DNA molecule. The results here suggest that the conjugation mechanisms of linear versus circular DNA in Streptomyces spp. are inherently different and/or that efficient transfer of linear DNA requires additional components.

    • ↵†Present address: Department of Chemistry, Rice University, Houston, TX 77005, USA.

    Edited by: C. W. Chen

    INTRODUCTION

    Mycelial actinomycetes such as Streptomyces spp. produce antibiotics and form spores as part of a complex developmental life cycle. These bacteria are further distinguished by a conjugation mechanism that is distinct from other eubacteria. Streptomyces circular plasmids encode transfer-essential membrane proteins (TraB proteins) which appear to be localized to the hyphal tips of mycelia (Reuther et al., 2006). TraB transfer proteins are related to the SpoIIIE/FtsK family of molecular motor proteins, which mediate translocation of double-stranded DNA across intracellular septa in other bacteria (Errington et al., 2001). Several lines of evidence imply that circular plasmids are transferred intercellularly as unprocessed double-stranded molecules during conjugation in Streptomyces spp. (Ducote & Pettis, 2006; Ducote et al., 2000; Possoz et al., 2001; Reuther et al., 2006). A potential mechanism of plasmid transfer, which invokes the DNA pumping mechanism of the Bacillus subtilis SpoIIIE protein (Errington et al., 2001), suggests that donor and recipient mycelial tips fuse and TraB protein, which may form a channel in the donor membrane, interacts non-covalently with double-stranded plasmid molecules to pump them into recipients by using ATP hydrolysis (Grohmann et al., 2003; Reuther et al., 2006).

    Circular Streptomyces plasmids also carry a cis-acting transfer locus (clt), which enhances the frequency of plasmid transfer significantly but, unlike TraB, is not required for transfer to occur. Addition of the clt locus of Streptomyces lividans pIJ101 to a transfer-defective version of either pIJ101 or an unrelated circular replicon resulted in a two to three order of magnitude increase in transfer, as mediated by the TraB-encoding pIJ101 tra gene (Ducote et al., 2000; Pettis & Cohen, 1994). The fact that clt− replicons were observed to transfer, albeit at lower frequencies than their clt+ derivatives, suggested that apparently unrelated sequences can partially complement the clt transfer function (Pettis & Cohen, 1994). Meanwhile, clt plasmid loci appear to be dispensable for plasmid-mediated (i.e. TraB-mediated) chromosome mobilization during mating (Pettis & Cohen, 1994). While the exact role of clt in plasmid transfer remains unclear, a direct interaction with TraB seems likely (Reuther et al., 2006).

    Unlike most other characterized eubacteria, Streptomyces spp. also possess linear chromosomes and linear plasmids. Streptomyces linear plasmids have many features in common with host chromosomes such as telomeres consisting of terminal inverted repeats (TIRs) and terminal protein (TP) covalently attached at the 5′ ends (Chen et al., 2002). In the case of SLP2, a 50 kb linear plasmid isolated from S. lividans (Chen et al., 2002; Huang et al., 2003), its right hand end is actually identical to the first 15.4 kb of the TIR of the S. lividans chromosome (Bey et al., 2000). Both linear chromosomes and plasmids also replicate divergently from centrally located origins of replication (Chang & Cohen, 1994; Chang et al., 1996; Musialowski et al., 1994) and can exist in a circular form (Chen et al., 2002; Shiffman & Cohen, 1992). Circularization can occur through artificial experimental manipulations (Lin & Chen, 1997; Lin et al., 1993; Shiffman & Cohen, 1992; Volff et al., 1997) or through natural genetic instability (Chen et al., 2002; Volff & Altenbuchner, 1998).

    Relatively less is known regarding conjugation of Streptomyces linear plasmids. Some linear plasmids are known to encode a SpoIIIE/FtsK homologue, which in the case of SLP2 promotes efficient conjugal transfer of not only linear SLP2 but also its circularized derivatives (Xu et al., 2006). For the transfer of linear plasmids in their natural configuration, an end-first model has been proposed where transfer is initiated from a TP-capped end (Chen, 1996). While the existence of clt loci for linear plasmids remains unknown, additional functions potentially unique to the transfer of linear molecules have been identified (Bentley et al., 2004; Huang et al., 2003; Xu et al., 2006). These include ttrA, a putative helicase gene which is present in a single copy near the end of both SLP2 and the S. lividans chromosome (Bey et al., 2000; Huang et al., 2003). Previously, the ttrA gene of SLP2 was shown to be important for SLP2 transfer, while both the SLP2 and chromosomal copies of ttrA were found to be required for SLP2-mediated mobilization of the S. lividans chromosome during mating (Huang et al., 2003).

    To investigate further the requirements for conjugation of circular versus linear DNA in Streptomyces spp., we tested whether the tra gene of circular plasmid pIJ101 can mediate transfer of linear versus circular versions of the same plasmid. Interestingly, it was found that linear versions transferred at exceedingly low frequencies and could only be recovered in a circularized form in recipients. Such circularization appeared to occur prior to transfer in donor cells. These same linear plasmids could be transferred, however, at much greater frequencies and in linear form by the SLP2 transfer locus. The data show that functions that were satisfactory for efficient transfer of circular DNA in Streptomyces bacteria were insufficient for effective transfer of linear DNA molecules, a result which strengthens the notion that significant differences exist between how circular and linear DNA molecules are transferred in these bacteria.

    METHODS

    Bacterial strains and plasmids.

    Escherichia coli and S. lividans strains, as well as plasmids used in this study, are listed in Table 1⇓. Strains JW1, JW10, JW11 and JW50 were confirmed for chromosomal integration of pSCON331 by Western blotting of KorA (Tai & Cohen, 1994) and Tra (Pettis & Cohen, 1996) in cell extracts as previously described. For the construction of pSCON335, the aadA gene was amplified using PCR conditions as described by Pettis et al. (2001), using primers 5aadARI (5′-TTTTTGAATTCCGAACGCAGCGGTGGTA-3′) and 3aadAMluI (5′-TTTTTACGC GTTATGTGCTTAGTGCATC-3′), with pSCON75 as template. The 1 kb product was digested with EcoRI and MluI and ligated to the 4.8 kb EcoRI–MluI fragment of pQC179 to create pSCON334. The 6.2 kb MluI fragment of pQC179 was then ligated, in the correct orientation, to pSCON334 at its MluI site to create pSCON335.

    Table 1.

    Bacterial strains and plasmids used in study

    General molecular biological and genetic methods.

    Cloning was achieved according to standard procedures described in Sambrook et al. (1989). Restriction enzymes and T4 ligase were purchased from New England Biolabs. A GeneClean Spin kit (MP Biomedicals) was used to purify DNA fragments after enzymic reactions. Plasmid DNA from S. lividans was isolated as described previously (Pettis & Cohen, 2000). Chemical transformation of E. coli strains cultured in LB broth and on LB agar was performed as described by Sambrook et al. (1989). Selection for plasmids in E. coli was achieved using, where appropriate, 50 μg ampicillin ml−1 and 200 μg apramycin ml−1. Protoplast transformation of S. lividans was performed using R5 agar according to the method described by Kieser et al. (2000). Transformant colonies of S. lividans were excised and patched onto Streptomyces ipomoeae growth agar (Clark et al., 1998) as previously described (Pettis et al., 2001). Liquid cultures of S. lividans were grown in yeast extract-malt extract (YEME) medium (Kieser et al., 2000) at 30 °C. Antibiotics used for appropriate selection of strains or plasmids in S. lividans were included at the following concentrations: 50 μg thiostrepton ml−1 in solid media, or 5 μg thiostrepton ml−1 in liquid media; 200 μg hygromycin ml−1 in solid media, or 20 μg hygromycin ml−1 in liquid media; and 400 μg spectinomycin ml−1, 30 μg streptomycin ml−1 and 50 μg apramycin ml−1 in both solid and liquid media. PCR sequencing reactions were typically performed as described by Qin & Cohen (1998) using primer tsr1 (5′-GCGGATGGCTGCCCTGAC-3′), mini-preparations of plasmid DNA from transconjugants as template and a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). For detection of circularized derivatives of pSCON223L in donor cells, PCR was performed using primers tsr1 and melC2 (5′-GTCCTCGCCGACGGGCGGAAG-3′), relevant genomic DNA as template and Phusion DNA Polymerase (New England Biolabs). PCR here was performed with an initial temperature of 98 °C for 10 s followed by 35 cycles consisting of 98 °C for 1 min, 55 °C for 15 s and 72 °C for 20 s, plus a final extension step of 72 °C for 10 min. Following electrophoresis, the major 0.4 kb PCR product was excised from the gel, A-tailed using Taq DNA polymerase (New England Biolabs), and the resulting fragment was cloned into pGEM-T easy vector (Promega). The insert was then sequenced using universal SP6 and T7 primers (New England Biolabs) and the method described above.

    Mating assays were performed and quantified as described previously (Pettis & Cohen, 1994). The data presented represent means from at least four independent trials. Upon transformation of S. lividans strains using covalently closed circular (CCC) pQC179 or its derivatives, either circular or linear DNA molecules were recovered in transformants, as has been described previously (Qin & Cohen, 1998). Identification of circular versus linear forms of plasmids, in donor or transconjugant cells, was performed by exonuclease III and λ exonuclease analyses of genomic DNA as described in the next Methods section.

    Isolation and analysis of Streptomyces genomic DNA.

    Preparation of total DNA from 10 ml liquid cultures was performed based on Qin & Cohen (1998) with modifications. Briefly, following lysozyme treatment and addition of proteinase K and SDS as described (Qin & Cohen, 1998), samples were incubated at 37 °C until lysis was complete. Following phenol/chloroform (50 : 50, v/v) extraction, 50 μg RNase A ml−1 was added and the mixture was incubated at 37 °C for 30 min. Phenol/chloroform extraction was repeated several times until no more interphase appeared. Then 0.1 vols 3 M sodium acetate (pH 5.2) and 0.6 vols 2-propanol were added and, after gentle mixing, the visible pellet of DNA was removed, washed with 70 % ethanol, and finally dissolved in 500 μl of TE. The configuration (linear or circular) of plasmids in total DNA was determined by digestions of exonuclease III and λ exonuclease (New England Biolabs) as described by Qin et al. (2003), followed by agarose gel electrophoresis. Digital imaging of finished gels was performed by using an Eagle Eye II still video system (Stratagene).

    Pulsed-field gel electrophoresis and Southern blotting.

    PFGE of undigested genomic DNA was performed in a CHEF DR III system (Bio-Rad) as described by Kers et al. (2005) except using a 6 V cm−1 voltage gradient and switch times ramped from 1 to 12 s at 12 °C for 15 h. The low-range PFG marker (New England Biolabs) was used for size markers. Following electrophoresis, gels were stained with ethidium bromide and digital imaging was performed as described earlier. Southern blotting was performed as previously described (Brasch et al., 1993; Grau et al., 2008) using a Vacugene XL vacuum blotting system as per the manufacturer's instructions. The 1.8 kb EcoRI-MluI fragment containing the tsr gene and the 1.0 kb EcoRI–MluI fragment containing the aadA gene, of pQC179 and pSCON335, respectively, served as templates for the generation of radioactive probes as described by Grau et al. (2008). Autoradiography of the radioactive blots was performed with Hyperfilm MP (Amersham Pharmacia Biotech) at −70 °C.

    RESULTS

    Circular, but not linear, versions of the same replicon are transferred efficiently from S. lividans strain TK23.42

    To investigate whether the pIJ101 conjugation system can mediate efficient transfer of a linear replicon, the clt locus of pIJ101 was inserted into the heterologous thiostrepton-resistant plasmid pQC179 (Qin et al., 2003), which contains the replication and stability region of Streptomyces linear plasmid pSLA2 flanked by cloned copies of the S. lividans chromosomal telomere sequences. Like pQC179, the resulting clt+ derivative pSCON223 can exist as either a circular or linear molecule in Streptomyces as shown in Fig. 1(a, b)⇓. The transfer frequencies of circular (C) and linear (L) versions of both pQC179 and pSCON223 were then determined by a previously described mating assay (Pettis & Cohen, 1994, 1996) involving the S. lividans donor strain TK23.42, which contains a chromosomally inserted copy of the pIJ101 tra gene and its regulation gene korA (Stein et al., 1989), and a TK23 recipient strain containing the transfer-defective hygromycin-resistant plasmid pHYG1.

    Figure image not available in archive
    Fig. 1.

    Physical and genetic maps of pSCON223C (a) pSCON223L (b) and pSCON346L (c). All plasmids are derived from pQC179 (Qin et al., 2003). Telomere sequences are derived from the S. lividans chromosome (Chr) and are represented by the portion of each plasmid map extending from the tip of the arrowhead to the nearby small horizontal line. Each TP is designated by a filled circle. The position and orientation of the tsr1 and melc 2 primers are indicated by the small arrows in (a). clt, minimal cis-acting locus of transfer derived from S. lividans plasmid pIJ101; tsr, thiostrepton-resistance gene; aadA, spectinomycin/streptomycin adenylyltransferase gene; melC, melanin biosynthesis gene; pSLA2, replication and stability region of Streptomyces rochei linear plasmid pSLA2; pSP72, replication and stability region of the E. coli cloning vector pSP72.

    Surprisingly, as shown in Table 2⇓, pQC179C transferred at a frequency nearly identical to that seen for its clt+ derivative pSCON223C. This result, which is discussed in more detail later, suggested that some sequence within pQC179 can complement the pIJ101 clt transfer function. Aside from the unexpectedly high transfer rate of pQC179C, the striking result for these initial matings was the exceptionally low transfer frequencies of pQC179L and pSCON223L from strain TK23.42. The three order of magnitude reduction in transfer of the linear versions of pQC179/pSCON223 versus their circular counterparts suggested that the pIJ101 TraB function encoded by strain TK23.42 was unable to promote efficient transfer of these linear plasmids. To verify that the observed transfer of pQC179C and pSCON223C from TK23.42 was mediated by pIJ101 TraB, transmission of these circular plasmids from the parental strain TK23, which lacks the pIJ101 tra and korA genes, was also determined; as expected, no transfer of pQC179C and pSCON223C was seen from strain TK23 (Table 2⇓).

    Table 2.

    Transfer frequencies of linear versus circular pQC179 and pSCON223 replicons from various S. lividans donor strains

    Using the indicated S. lividans donor strains and plasmids and TK23(pHYG1) as the recipient, matings were performed and quantified as described by Pettis & Cohen (1994).

    To determine whether pQC179L or pSCON223L could be recovered in linear form following mating from TK23.42, total DNA was isolated from individual transconjugants and treated with exonuclease III or λ exonuclease. Following such treatment, linear plasmids display a single λ exonuclease-resistant, exonuclease III-sensitive band on agarose gels (Hirochika & Sakaguchi, 1982); meanwhile circular molecules were found here to show mainly an exonuclease III-resistant, λ exonuclease-resistant CCC form. As indicated by the representative profile in Fig. 2(a)⇓, every transconjugant examined showed evidence of the CCC form of the donor plasmid with no evidence of the linear form that was present in donor cells; thus, pQC179L or pSCON223L derivatives isolated from transconjugants always appeared to be in the circular configuration.

    Figure image not available in archive
    Fig. 2.

    Exonuclease III and λ exonuclease analyses of total DNA from transconjugants. Mating assays were performed as described by Pettis & Cohen (1994) and total DNA was isolated from individual transconjugant isolates as indicated in Methods. For each transconjugant, a 2 μg aliquot of DNA undigested (U), treated with either exonuclease III (III) or λ exonuclease (λ) as described by Qin et al. (2003) was electrophoresed on a 0.7 % agarose gel containing ethidium bromide at 45 V for 4 h. Gels were then visualized under UV light and a digital image was recorded. (a) The donor strain was TK23.42(pSCON223L) and the recipient was TK23(pHYG1). Shown is an example of pSCON223 observed as circular in transconjugants (CT). Plasmids pSCON223L in the donor (LD), pSCON223C in the donor (CD) and undigested pHYG1 isolated from recipient cells were included as controls. Pertinent CCC and OC plasmid forms are as indicated. The 1 kb plus DNA ladder (Invitrogen) was used as a size marker (M). (b) The donor strain was JW1(pQC179L) or JW1(pSCON223L) and the recipient was TK23(pHYG1). Representative transconjugant isolates derived from the mating involving JW1(pQC179L) were 179.1 and 179.2, while isolates derived from the mating involving JW1(pSCON223L) were 223.3 and 223.4. Based on subsequent PFGE analysis (see text for details), the slowest migrating pQC179 band for isolate 179.2 is predicted to be a mixture of OC and randomly linearized versions of the circularized pQC179 plasmid (OC species should be sensitive to exonuclease III and resistant to λ exonuclease). Undigested plasmid pHYG1 was included as a control. The 1 kb plus DNA ladder was used as a size marker (M).

    As the presence of pHYG1 in transconjugants made interpretation of these results somewhat tenuous, transconjugant plasmid DNA isolated by the alkaline lysis method was sequenced using a primer (designated tsr1; see Methods), which is specific for the 3′ end of the tsr gene and oriented toward the nearby telomere sequence on pQC179/pSCON223 plasmids as shown in Fig. 1(a)⇑. Of the 15 individual transconjugants analysed (i.e. twelve from nine independent pQC179L or pSCON223L matings and three from three separate pQC179C or pSCON223C matings), DNA sequencing extended into the melC gene in every case (data not shown), which indicated that the linear molecules had circularized (Fig. 1⇑). Additionally, of the 12 plasmids derived from either pSCON223L or pQC179L, one retained intact telomeres, seven others retained a partial telomere region ranging from ∼50 bp to ∼150 bp, while the remaining four had lost all telomere sequences. Total loss of telomeres was also observed for two of the plasmids derived from either pQC179C or pSCON223C, while the other one remained intact.

    pSCON223L does transfer in linear form from S. lividans strain JW1

    As compared to the results for strain TK23.42, transfer frequencies for pQC179L and pSCON223L dramatically improved when the donor was S. lividans JW1 (Table 2⇑), which in addition to the pIJ101 tra and korA genes, also contains the autonomous linear plasmid SLP2. Transconjugants from independent matings were examined as before by isolation of total DNA and treatment with exonuclease III and λ exonuclease. For the pQC179L matings, two out of eight transconjugants appeared to contain circularized versions of pQC179, while neither circular nor linear forms of pQC179 were detected for the other six transconjugants. The latter effect could result from recombination between the identical terminal sequences of pQC179 and SLP2 such that a composite molecule of some form would be created. For the pSCON223L matings, six out of eight transconjugants contained linear pSCON223L, while the other two appeared to contain a circular version of this plasmid. Fig. 2(b)⇑ shows representative transconjugant isolates from this analysis: isolate 179.1, which showed neither circular nor linear forms of pQC179; isolate 179.2, which appeared to contain pQC179 in a circular form; isolate 223.3, which appeared to contain pSCON223L; and isolate 223.4, which also showed evidence of pSCON223L, but at an apparently lower copy number than 223.3.

    To further examine the status of plasmids pQC179 or pSCON223 in these representative transconjugants, their undigested total DNA was analysed by PFGE (Fig. 3a⇓) and subsequent Southern blotting using a pQC179/pSCON223-specific tsr gene probe (Fig. 3b⇓). All transconjugant isolates appeared to contain SLP2 or some derivative thereof (Fig. 3a⇓). Additionally, autonomous pQC179 plasmids were confirmed to be absent in isolate 179.1, though pQC179-specific sequences were associated with SLP2 (Fig. 3b⇓, lane 1). Since there was no evidence of recombination between pQC179L and SLP2 in the original non-mating donor strain, it is possible this event occurred during transfer. This scenario would be similar to a previous study in which homologous recombination between the conjugative Streptomyces plasmid SCP2* and a non-conjugative replicon during mating led to the formation of a transmissible cointegrate structure (Xiao et al., 1994). Meanwhile, a circular version of pQC179 was confirmed here for isolate 179.2 (Fig. 3b⇓, lane 2), while the presence of pSCON223L was confirmed for both isolates 223.3 and 223.4 (Fig. 3b⇓, lanes 3 and 4 respectively). A larger apparently cointegrate SLP2/pSCON223 molecule was also identified in the latter transconjugant. Taken together, the results suggest that pSCON223L (and probably pQC179L also) can be transferred in linear form from strain JW1 at frequencies approximately equivalent to that seen for pSCON223C (and pQC179C) from strain TK23.42. Following transfer from JW1, pSCON223L can then be recovered as an autonomous linear molecule from transconjugants (e.g. isolates 223.3 and 223.4).

    Figure image not available in archive
    Fig. 3.

    PFGE and Southern blot analysis to determine configuration of plasmid pQC179 or pSCON223 following transfer from strain JW1. (a) PFGE analysis. Following mating of the donor strain JW1(pQC179L) or JW1(pSCON223L) and the recipient TK23(pHYG1), agarose plugs containing mycelia of individual transconjugant isolates were prepared and PFGE of undigested genomic DNA was performed all as described by Kers et al. (2005), except using the run parameters outlined in Methods. JW1(pQC179L) and JW1(pSCON223L) were included as linear donor (LD) controls. The low-range PFG marker (New England Biolabs) was used as a size marker (M). The migration positions of linear plasmids SLP2, pQC179L and pSCON223L are as indicated. Lanes: 1, isolate 179.1; 2, isolate 179.2; 3, isolate 223.3; 4, isolate 223.4. (b) Southern blot analysis. The gel in (a) was subjected to Southern blotting using a 32P-labelled tsr gene probe and resulting bands were visualized by autoradiography. The circularized version of pQC179 in isolate 179.2 appears to be represented by CCC (fastest migrating), OC (slowest migrating) and randomly linearized (i.e. not true pQC179L) forms. The relative positions of the 48.5 kb and 9.42 kb bands of the size marker are shown.

    Linear pQC179 derivatives do transfer in linear form from strains containing only SLP2

    The enhanced transfer rates of pQC179L and pSCON223L from strain JW1 could be the result of the pIJ101 TraB function working in concert with one or more ancillary functions provided by SLP2 (e.g. ttrA) and/or simply by mobilization of these plasmids via the SLP2 conjugation system itself. In either case, given that the ttrA gene was previously shown to be important for efficient SLP2 transfer (Huang et al., 2003), its potential role in the transfer of linear pQC179/pSCON223 molecules from strain JW1 was next investigated. Specifically, transfer of the spectinomycin/streptomycin-resistant pQC179 derivative pSCON335L as well as its clt+ derivative pSCON346L (Fig. 1c⇑) was assessed from donor strains similar to JW1, except that they were deleted for the SLP2 ttrA gene (JW50), the chromosomal ttrA gene (JW10), or both (JW11). The clt+ derivative was included here since the 70-fold difference in transfer frequency of pSCON223L versus pQC179L from strain JW1 (Table 2⇑) raised the possibility that the presence of clt might be affecting transfer frequency. As shown in Table 3⇓, plasmids pSCON335L and pSCON346L transferred at high, yet comparable frequencies from strains JW10, JW11, and JW50, which indicated that neither the loss of ttrA nor the presence of clt appeared to affect transmission of these plasmids from these strains.

    Table 3.

    Transfer frequencies of pSCON335L and pSCON346L from ttrA defective strains

    Matings were performed and quantified as described for Table 2.

    To determine directly whether SLP2 conjugation functions alone can efficiently mobilize linear pQC179/pSCON223 plasmids, transfer of pSCON223L from S. lividans strain 1326 was measured. Strain 1326 is the parent of JW1, and contains SLP2 (as well as the uncharacterized conjugative element SLP3) (Kieser et al., 2000) but lacks chromosomal copies of the pIJ101 tra and korA genes. As shown in Table 4⇓, the transfer frequency of pSCON223L from strain 1326 was within a few fold of its transfer frequency from JW1 and was nevertheless approximately three orders of magnitude greater than its rate of transfer from strain TK23.42 (Table 2⇑). In an additional relevant mating, plasmid pSCON346L was found to transfer at a high frequency from donor strain 98-50 (Table 4⇓). This strain is the parent of JW50, and contains SLP2 with its ttrA gene deleted, but lacks SLP3 as well as chromosomal copies of the pIJ101 tra and korA genes. Upon examination of transconjugants from the latter mating by PFGE and Southern blotting using an aadA gene probe, linear pSCON346 molecules were readily identified (e.g. isolate 346.1) (Fig. 4a, b⇓, lanes 1).

    Figure image not available in archive
    Fig. 4.

    PFGE and Southern blot analysis to determine configuration of plasmid pSCON346 following transfer from strain 98-50. (a) PFGE analysis. Following mating of the donor strain 98-50(pSCON346L) and the recipient TK23(pHYG1), PFGE was performed on transconjugant isolates as described in the legend to Fig. 3. The low-range PFG marker (New England Biolabs) was used as a size marker (M). The migration positions of linear plasmids SLP2 and pSCON346L are as indicated. Lanes: 1, transconjugant isolate 346.1; 2, TK23(pHYG1). (b) Southern blotting analysis. The gel in (a) was subjected to Southern blotting using a 32P-labelled aadA gene probe and the resulting band was visualized by autoradiography. The relative position of the 9.42 kb band of the size marker is shown.

    Table 4.

    Transfer frequencies of pSCON223L and pSCON346L from SLP2-containing strains

    Matings were performed and quantified as described for Table 2.

    Evidence of circularization of pSCON223L in the donor strain TK23.42

    The results here for donor strain TK23.42 versus SLP2-containing donors suggested that TraB(pIJ101) is unable to effectively transfer linear plasmids and that relatively rare circularization of linear plasmids such as pSCON223L in donors may be a prerequisite for subsequent transfer by TraB(pIJ101). To examine whether such circularization indeed occurs, a TK23.42(pSCON223L) donor (Fig. 5a⇓) was plated either alone on the medium used for conjugations or mixed with TK23(pHYG1) recipient spores. Following one life cycle, spores from mating patches were quantified as usual in order to verify the characteristic low transfer frequency of pSCON223L from TK23.42 (data not shown); meanwhile, TK23.42(pSCON223L) spores collected from non-mating patches were inoculated into YEME broth containing thiostrepton, and genomic DNA was subsequently isolated from the resulting cultures and used as template in PCRs that included primers tsr1 and melC2 (Fig. 1a⇑). A major amplified product of ∼0.4 kb (Fig. 5b⇓), which was potentially indicative of circularization, was cloned and the inserts from multiple clones were sequenced. In every case, the sequencing results indicated that the amplified region corresponded to relevant portions of the tsr and melC genes, with no evidence of intervening telomere sequences (data not shown).

    Figure image not available in archive
    Fig. 5.

    Evidence of circularization of pSCON223L in TK23.42 donor cells. (a) Exonuclease III and λ exonuclease analysis of the TK23.42(pSCON223L) donor strain used in this experiment. Aliquots of genomic DNA from this donor were treated with exonuclease III (III) or λ exonuclease (λ), or were left untreated (U), and were then subjected to gel electrophoresis all as described in the legend to Fig. 2. The position of plasmid pSCON223L in the donor (LD) is indicated. (b) PCR amplification of a fragment indicative of circularization. Following growth for one life cycle under non-mating conditions, genomic DNA was isolated from broth cultures of the donor strain and was used as template in PCRs that included primers tsr1 and melC2 (Fig. 1a). PCR products were visualized by electrophoresis on 1 % agarose gels containing ethidium bromide. The 1 kb plus DNA ladder was used as a size marker (M) and the 0.4 kb fragment of the ladder is highlighted. The results in lanes 1 and 2 are derived from PCRs involving genomic DNA corresponding to two independent non-mating spore patches of the donor.

    As a separate test for circularization in donors, genomic DNA for TK23.42(pSCON223L) was treated extensively with exonuclease III to eliminate linear pSCON223L (as judged by gel electrophoresis) and this mixture was then used to transform strain TK23 to thiostrepton resistance. Of the seven transformants obtained by this procedure, four were analysed further by miniprep DNA isolation and subsequent gel electrophoresis of recovered plasmids. In all cases examined, the undigested plasmids showed multiple bands (data not shown), including a faster-migrating species that approximated the CCC form of pSCON223 that was always seen in transconjugants from matings involving the TK23.42 donor (e.g. Fig. 2a⇑).

    DISCUSSION

    For the S. lividans donor strain TK23.42, the chromosomally inserted tra gene in conjunction with the plasmid-borne clt locus, both of pIJ101 origin, has been shown to promote efficient transfer of both pIJ101 and non-cognate transfer-defective circular replicons (Ducote et al., 2000; Pettis & Cohen, 1994). All other putative functions required for efficient circular plasmid transfer from this strain must be provided by the host chromosome. Here, we have shown that under mating conditions resulting in efficient TraB-mediated transfer of circular pQC179 or pSCON223 from TK23.42, exceedingly low transfer rates of linear versions of these plasmids were seen, and transferred plasmids were then recovered only in circular form in transconjugants. In the latter case, such transfers may have been entirely dependent on the relatively rare circularization of linear plasmids observed in donors. Meanwhile, efficient transfer of pSCON223L or pSCON346L was seen from donor strains that included SLP2, and these plasmids were readily recoverable as linear molecules from transconjugants following mating. These data indicate that the minimum functions needed for efficient transfer of circular plasmids were insufficient for transfer of linear versions of these same plasmids. This implies that the mechanisms of circular versus linear DNA transfer in streptomycetes must differ in some substantial way(s).

    In a study of chromosomal recombinants following plasmid-mediated interspecies conjugation, it was concluded that linear plasmids mediate transfer of the host chromosome beginning at one end (with complete transfer usually occurring), while circular plasmids mediate transfer beginning from an internal position, with transfer extending to one end or the other only occasionally (Wang et al., 1999). Thus, while the presence of SpoIIIE/FtsK-related TraB functions for both circular and linear Streptomyces plasmids may indicate that the principle of dsDNA transfer is conserved, the exact mechanisms of circular versus linear plasmid-mediated transfer may differ in other respects (e.g. where the TraB translocase interacts initially with the DNA molecule to begin the transfer process). Another intriguing question remaining to be answered is whether a cis-acting function analogous to clt is required for efficient linear plasmid-mediated DNA transfer.

    Interestingly, conjugation systems of linear plasmids have been shown to promote transfer of circular plasmid replicons (Xu et al., 2006; Zotchev & Schrempf, 1994). Here, however, the reverse situation – conjugation functions of a circular plasmid mediating transfer of a linear plasmid replicon – was not observed. Aside from the potential mechanistic differences with respect to TraB and clt functions as discussed in the previous paragraph, an additional relevant factor may be that, for reasons still unclear, transfer of linear DNA in streptomycetes appears to require more functions than transfer of circular DNA molecules. Both linear chromosomes and linear plasmids (e.g. SLP2, SCP1) have been shown to encode ttrA helicase-like genes (Bentley et al., 2002; Huang et al., 2003), and efficient transfer of the S. lividans chromosome and plasmid SLP2 was previously shown to be dependent on ttrA (Huang et al., 2003). Although ttrA did not appear to be required for efficient transfer of the linear plasmids used in our study, SLP2 is known to encode additional functions (e.g. a lytic transglycosylase, two cell membrane proteins, and an ATP-binding protein) (Huang et al., 2003; Xu et al., 2006), which also appear to be involved in conjugal DNA transfer (Xu et al., 2006). Perhaps the presence of one or more of these functions was required for effective transfer of linear pQC179 derivative plasmids here.

    We speculate that the relatively rare circularization of linear plasmids observed here in donor cells is the result of some genetic instability associated with the linear molecules, a situation that is reminiscent of Streptomyces chromosomes themselves (Chen et al., 2002; Volff & Altenbuchner, 1998). Alignment of homologous TIR sequences present at the two ends of the linear plasmids (e.g. in a partially overlapping circular-like configuration) with subsequent single crossover between them could, for example, generate circular derivatives. The circularized derivatives formed would then be capable of being transferred intercellularly during mating by the pIJ101 transfer apparatus. It is not known to what extent the likely high copy number of the pQC179 derivatives used here (Qin et al., 2003) may have influenced their frequency of circularization and/or eventual transfer. Judging from our sequence analysis, subsequent deletion of remaining TIR sequences in these circularized derivatives probably also sometimes occurred, which is consistent with the deletion of telomeric sequences also observed for the circular plasmids pQC179C and pSCON223C. Overall, these various deletion events appear to resemble those which occurred following initial artificial circularization of the S. lividans chromosome (Chen et al., 2002; Lin & Chen, 1997).

    The unexpectedly high transfer frequency for pQC179C from strain TK23.42, which approximated the frequency seen for pSCON223C, indicated that certain unidentified sequences within pQC179C can provide near, or perhaps complete, complementation of the pIJ101 clt function. Various transfer-defective versions of pIJ101 or SCP2* plasmids lacking clt(pIJ101) have been examined using this same TraB(pIJ101) mating system and they typically transfer at frequencies two to three orders of magnitude below that seen for clt+ derivatives (Ducote et al., 2000; Pettis & Cohen, 1994; Pettis & Prakash, 1999). This includes composite replicons involving the entire E. coli-based cloning vector pSP72, which was also used in construction of pQC179C (Qin et al., 2003). It seems likely therefore that the complementing sequences reside within either the cloned pSLA2 replication and stability region or the cloned chromosomal telomeres on pQC179C.

    The minimal pIJ101 clt locus is composed of only 54 bp and consists of a single inverted repeat region along with three adjacent small direct repeats. Deletion analysis has demonstrated that sequences within the inverted repeat appear to be the most critical for clt function (Ducote et al., 2000). Additionally, evidence of intrinsic curvature within the locus raises the possibility that secondary structuring of clt sequences may be important for its function (Ducote et al., 2000). Although we found no sequences exactly matching clt within the composite pQC179C sequence, it is interesting to note that the inverted repeat region of clt does show some similarity to the palindromic TIR sequences of the S. lividans chromosomal telomeres (data not shown). Experiments are under way to determine the exact sequences present in pQC179C that appear to be complementing the pIJ101 clt transfer function.

    Acknowledgments

    We thank Zhongjun Qin and Stanley Cohen for plasmids and Carten Chen for strains. We are very grateful to Ronald Parry, who graciously provided laboratory space and supplies for certain confirmatory experiments. We also thank Xue Bai for preparation of the figures in this manuscript. J. W. was the recipient of an Economic Development Assistantship from the Louisiana State University Board of Regents.

    References

    1. Bentley, S. D., Chater, K. F., Cerdeno-Tarraga, A.-M., Challis, G. L., Thomson, N. R., James, K. D., Harris, D. E., Quall, M. A., Kieser, H. & other authors (2002). Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature 417, 141–147.
    2. Bentley, S. D., Brown, S., Murphy, L. D., Harris, D. E., Quail, M. A., Parkhill, J., Barrell, B. G., McCormick, J. R., Santamaria, R. I. & other authors (2004). SCP1, a 356 023 bp linear plasmid adapted to the ecology and developmental biology of its host, Streptomyces coelicolor A3(2). Mol Microbiol 51, 1615–1628.
    3. Bey, S. J., Tsou, M. F., Huang, C. H., Yang, C. C. & Chen, C. W. (2000). The homologous terminal sequence of the Streptomyces lividans chromosome and SLP2 plasmid. Microbiology 146, 911–922.
    4. Bierman, M., Logan, R., O'Brien, K., Seno, E. T., Rao, R. N. & Schoner, B. E. (1992). Plasmid cloning vectors for the conjugal transfer of DNA from Escherichia coli to Streptomyces spp. Gene 116, 43–49.
    5. Brasch, M. A., Pettis, G. S., Lee, S. C. & Cohen, S. N. (1993). Localization and nucleotide sequences of genes mediating site-specific recombination of the SLP1 element in Streptomyces lividans. J Bacteriol 175, 3067–3074.
    6. Chang, P.-C. & Cohen, S. N. (1994). Bidirectional replication from an internal origin in a linear Streptomyces plasmid. Science 265, 952–954.
    7. Chang, P.-C., Kim, E.-S. & Cohen, S. N. (1996). Streptomyces linear plasmids that contain a phage-like, centrally located, replication origin. Mol Microbiol 22, 789–800.
    8. Chen, C. W. (1996). Complications and implications of linear bacterial chromosomes. Trends Genet 12, 192–196.
    9. Chen, C. W., Huang, C.-H., Lee, H.-H., Tsai, H.-H. & Kirby, R. (2002). Once the circle has been broken: dynamics and evolution of Streptomyces chromosomes. Trends Genet 18, 522–529.
    10. Clark, C. A., Chen, C., Ward-Rainey, N. & Pettis, G. S. (1998). Diversity within Streptomyces ipomoeae based on inhibitory interactions, rep-PCR, and plasmid profiles. Phytopathology 88, 1179–1186.
    11. Ducote, M. J. & Pettis, G. S. (2006). An in vivo assay for conjugation-mediated recombination yields novel results for Streptomyces plasmid pIJ101. Plasmid 55, 242–248.
    12. Ducote, M. J., Prakash, S. & Pettis, G. S. (2000). Minimal and contributing sequence determinants of the cis-acting locus of transfer (clt) of streptomycete plasmid pIJ101 occur within an intrinsically curved plasmid region. J Bacteriol 182, 6834–6841.
    13. Errington, J., Bath, J. & Wu, L. J. (2001). DNA transport in bacteria. Nat Rev Mol Cell Biol 2, 538–544.
    14. Grau, B. L., Henk, M. C., Garrison, K. L., Olivier, B. J., Schulz, R. M., O'Reilly, K. L. & Pettis, G. S. (2008). Further characterization of Vibrio vulnificus rugose variants and identification of a capsular and rugose exopolysaccharide gene cluster. Infect Immun 76, 1485–1497.
    15. Grohmann, E., Muth, G. & Espinosa, M. (2003). Conjugative plasmid transfer in gram-positive bacteria. Microbiol Mol Biol Rev 67, 277–301.
    16. Hirochika, H. & Sakaguchi, K. (1982). Analysis of linear plasmids isolated from Streptomyces: association of protein with the ends of the plasmid DNA. Plasmid 7, 59–65.
    17. Huang, C. H., Chen, C. Y., Tsai, H. H., Chen, C., Lin, Y. S. & Chen, C. W. (2003). Linear plasmid SLP2 of Streptomyces lividans is a composite replicon. Mol Microbiol 47, 1563–1576.
    18. Kendall, K. J. & Cohen, S. N. (1987). Plasmid transfer in Streptomyces lividans: identification of a kil-kor system associated with the transfer region of pIJ101. J Bacteriol 169, 4177–4183.
    19. Kers, J. A., Cameron, K. D., Joshi, M., Bukhalid, R. A., Morello, J. & Loria, R. (2005). A large, mobile pathogenicity island confers plant pathogenicity on Streptomyces species. Mol Microbiol 55, 1025–1033.
    20. Kieser, T., Bibb, M. J., Buttner, M. J., Chater, K. F. & Hopwood, D. A. (2000). Practical Streptomyces Genetics. Norwich, UK: John Innes Foundation.
    21. Lin, Y.-S. & Chen, C. W. (1997). Instability of artificially circularized chromosomes of Streptomyces lividans. Mol Microbiol 26, 709–719.
    22. Lin, Y.-S., Kieser, H. M., Hopwood, D. A. & Chen, C. W. (1993). The chromosomal DNA of Streptomyces lividans 66 is linear. Mol Microbiol 10, 923–933.
    23. Musialowski, M. S., Flett, F., Scott, G. B., Hobbs, G., Smith, C. P. & Oliver, S. G. (1994). Functional evidence that the principal DNA replication origin of the Streptomyces coelicolor chromosome is close to the dnaA-gyrB region. J Bacteriol 176, 5123–5125.
    24. Pettis, G. S. & Cohen, S. N. (1994). Transfer of the pIJ101 plasmid in Streptomyces lividans requires a cis-acting function dispensable for chromosomal gene transfer. Mol Microbiol 13, 955–964.
    25. Pettis, G. S. & Cohen, S. N. (1996). Plasmid transfer and expression of the transfer (tra) gene product of plasmid pIJ101 are temporally regulated during the Streptomyces lividans life cycle. Mol Microbiol 19, 1127–1135.
    26. Pettis, G. S. & Cohen, S. N. (2000). Mutational analysis of the tra locus of the broad-host-range Streptomyces plasmid pIJ101. J Bacteriol 182, 4500–4504.
    27. Pettis, G. S. & Prakash, S. (1999). Complementation of conjugation functions of Streptomyces lividans plasmid pIJ101 by the related Streptomyces plasmid pSB24.2. J Bacteriol 181, 4680–4685.
    28. Pettis, G. S., Ward, N. & Schully, K. L. (2001). Expression characteristics of the transfer-related kilB gene product of Streptomyces plasmid pIJ101: implications for the plasmid spread function. J Bacteriol 183, 1339–1345.
    29. Possoz, C., Ribard, C., Gagnat, J., Pernodet, J. L. & Guerineau, M. (2001). The integrative element pSAM2 from Streptomyces: kinetics and mode of conjugal transfer. Mol Microbiol 42, 159–166.
    30. Qin, Z. & Cohen, S. N. (1998). Replication at the telomeres of the Streptomyces linear plasmid pSLA2. Mol Microbiol 28, 893–903.
    31. Qin, Z., Shen, M. & Cohen, S. N. (2003). Identification and characterization of a pSLA2 plasmid locus required for linear DNA replication and circular plasmid stable inheritance in Streptomyces lividans. J Bacteriol 185, 6575–6582.
    32. Reuther, J., Gekeler, C., Tiffert, Y., Wohlleben, W. & Muth, G. (2006). Unique conjugation mechanism in mycelial streptomycetes: a DNA-binding ATPase translocates unprocessed plasmid DNA at the hyphal tip. Mol Microbiol 61, 436–446.
    33. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
    34. Schully, K. L. (2005). Regulatory intricacies that confer temporal control of conjugation and antibiotic production functions in Streptomyces: new variations on old themes. PhD dissertation, Louisiana State University.
    35. Shiffman, D. & Cohen, S. N. (1992). Reconstruction of a Streptomyces linear replicon from separately cloned DNA fragments: existence of a cryptic origin of circular replication within the linear plasmid. Proc Natl Acad Sci U S A 89, 6129–6133.
    36. Stein, D. S., Kendall, K. J. & Cohen, S. N. (1989). Identification and analysis of transcriptional regulatory signals for the kil and kor loci of Streptomyces plasmid pIJ101. J Bacteriol 171, 5768–5775.
    37. Tai, J. T.-N. & Cohen, S. N. (1994). Mutations that affect regulation of the korB gene of Streptomyces lividans plasmid pIJ101 alter plasmid transmission. Mol Microbiol 12, 31–39.
    38. Volff, J. N. & Altenbuchner, J. (1998). Genetic instability of the Streptomyces chromosome. Mol Microbiol 27, 239–246.
    39. Volff, J. N., Viell, P. & Altenbuchner, J. (1997). Artificial circularization of the chromosome with concomitant deletion of its terminal inverted repeats enhances genetic instability and genome rearrangement in Streptomyces lividans. Mol Gen Genet 253, 753–760.
    40. Wang, S. J., Chang, H. M., Lin, Y. S., Huang, C. H. & Chen, C. W. (1999). Streptomyces genomes: circular genetic maps from the linear chromosomes. Microbiology 145, 2209–2220.
    41. Xiao, J., Melton, R. E. & Kieser, T. (1994). High-frequency homologous plasmid-plasmid recombination coupled with conjugation of plasmid SCP2* in Streptomyces. Mol Microbiol 14, 547–555.
    42. Xu, M.-X., Zhu, Y.-M., Shen, M.-J., Jiang, W.-H., Zhao, G.-P. & Qin, Z.-J. (2006). Characterization of the essential gene components for conjugal transfer of Streptomyces lividans linear plasmid SLP2. Prog Biochem Biophys 33, 986–993.
    43. Zotchev, S. B. & Schrempf, H. (1994). The linear Streptomyces plasmid pBL1: analyses of transfer functions. Mol Gen Genet 242, 374–382.