Genome engineering and direct cloning of antibiotic gene clusters via phage ϕBT1 integrase-mediated site-specific recombination in Streptomyces

Several strategies have been used to clone large DNA fragments directly from bacterial genome. Most of these approaches are based on different site-specific recombination systems consisting of a specialized recombinase and its target sites. In this study, a novel strategy based on phage ϕBT1 integrase-mediated site-specific recombination was developed, and used for simultaneous Streptomyces genome engineering and cloning of antibiotic gene clusters. This method has been proved successful for the cloning of actinorhodin gene cluster from Streptomyces coelicolor M145, napsamycin gene cluster and daptomycin gene cluster from Streptomyces roseosporus NRRL 15998 at a frequency higher than 80%. Furthermore, the system could be used to increase the titer of antibiotics as we demonstrated with actinorhodin and daptomycin, and it will be broadly applicable in many Streptomyces.

S treptomyces are high-GC Gram-positive bacteria well known for their ability to produce a wide variety of medically and agriculturally useful antibiotics and related compounds 1 . Genes responsible for the biosynthesis of a specific secondary metabolite are usually arranged in clusters that vary in size from a few to over 100 kb 2 . To gain insight into the biosynthesis and regulation of antibiotics in Streptomyces, it is of great importance to clone their gene clusters. Recently, various approaches have been developed to clone gene clusters directly from bacterial genomic DNA. These methods include RecET-mediated linear-plus-linear homologous recombination (LLHR) 3 , oriT-directed capture system 4 and transformation-associated recombination (TAR) 5 . The RecET-mediated LLHR was successful in cloning gene clusters (10 to 52 kb in length) from the genome of Photorhabdus luminescens into expression vectors in Escherichia coli 3 . The oriT-directed capture system has been used to clone regions up to 140 kb from the genome of Burkholderia pseudomallei 6 and 200 kb from megaplasmid of Sinorhizobium meliloti 4 . However, the use of this system was limited to Gram-negative bacteria that can be established as conjugation donors 6 . Taking advantage of the natural in vivo homologous recombination of Saccharomyces cerevisiae, TAR cloning strategy was used to capture a 21.3 kb enterocin gene cluster from Salinispora pacifica CNT-150 7 and a 67 kb taromycin A biosynthetic gene cluster from Saccharomonospora sp. CNQ-490 8 .
The ability to delete large genomic fragments within Streptomyces genome is of great interest for genetic manipulations of Streptomyces. Several strategies have been developed for a number of bacteria. Some methods are based on the meganuclease I-SceI system which involves the meganuclease I-SceI of Saccharomyces cerevisiae and its 18 bp recognition sequence 9,10 . Many of them are based on site-specific recombination systems consisting of a specialized recombinase and its target sites. Nearly all site-specific recombinases fall into two families, the tyrosine recombinases and the serine recombinases 11 . The recombination systems of the tyrosine recombinase family include Cre/loxP from the P1 phage 12 , Dre/rox from the P1-like transducing phage D6 13 and the Flp/FRT from yeast 14 . The Cre, Dre and Flp proteins are the tyrosine recombinases which catalyze reciprocal site-specific recombination of DNA at loxP, rox and FRT sites, respectively. Integrases (Int) from Streptomyces temperate phage wC31 and wBT1 belong to serine recombinase family. They catalyze site-specific recombination of the phage attachment site (attP) with the bacterial attachment site (attB), resulting in the formation of two hybrid sites (attL and attR) 15,16 . Both wC31 and wBT1 attP-int loci have been used to construct versatile vectors which can integrate into different attB sites in Streptomyces 15,16 . To increase the diversity of attP-attB pair of wBT1, 15 mutated attP-attB pairs (attP 01 -attB 01 R attP 15 -attB 15 ) were generated by PCR mutagenesis of the central dinucleotide sequence of attB and attP 17 . The Cre/loxP system was successfully used for the deletion of large fragments in Magnetospirillum gryphiswaldense and several Streptomyces species [18][19][20] . However, the use of wC31 and wBT1 integrase in this aspect has not been exploited.
We devised a novel strategy for Streptomyces genome engineering and cloning of antibiotic gene clusters. This method is based on phage wBT1 attP-attB-int system and requires two single crossovers for targeted integration of mutated attB and attP into the recipient chromosome. Using the system, we easily cloned 25 kb fragment containing actinorhodin (act) gene cluster from S. coelicolor M145, 45 kb fragment containing napsamycin (nap) gene cluster and 157 kb fragment containing daptomycin (dap) gene cluster from S. roseosporus NRRL 15998. In addition, this method could be used to improve the titer of antibiotics by increasing copy numbers of antibiotic gene clusters.

Results
Construction of pUC119-and pKC1139-based plasmids. Our strategy used in this study requires both homologous and sitespecific recombinations. The homologous recombinations were used for targeted integration of the mutated attB and attP into Streptomyces chromosome, while the wBT1 integrase-mediated site-specific recombination was employed to excise targeted region of interest from the chromosome (Fig. 1). The mutated attB and attP sites were chosen to avoid site specific recombination with the endogenous attB site in Streptomyces genome and consequently undesirable DNA rearrangements. Sites of attB 6 and attP 6 were randomly chosen from the 15 mutated attP-attB pairs. For the integration of attB 6 into Streptomyces chromosome, pUC119-based suicide plasmids (pSV::attB 6 -act, pSV::attB 6 -nap and pSV::attB 6dap) were constructed. These plasmids are derivatives of pUC119 containing the kanamycin-resistance gene (neo), the origin of transfer (oriT) from plasmid RK2 (for the intergeneric conjugation between E. coli and Streptomyces), attB 6 and a 2.0 kb homologous region flanking 59 end of the targeted regions (Fig. S1a). We also constructed pKC1139-based plasmids (pKC1139::attP 6 -act, pKC1139:: attP 6 -nap and pKC1139::attP 6 -dap) for the integration of attP 6 into Streptomyces chromosome. These plasmids are derivatives of pKC1139 containing attP 6 and a 2.0 kb homologous region flanking 39 end of the targeted regions (Fig. S1b).
Cloning of act gene cluster from S. coelicolor M145. To test this strategy, we first chose to clone the well-studied act gene cluster (SCO5070-SCO5092) from S. coelicolor M145. For this purpose, pSV::attB 6 -act and pKC1139::attP 6 -act were introduced into the recipient chromosomes via single-crossover homologous recombination to obtain double-cointegrate strain Sco-actB 6 P 6 ( Fig. 1). Further introduction of pIJ10500 (an integrative plasmid containing the wBT1 integrase gene) into Sco-actB 6 P 6 allowed subsequent excision of 23 kb act gene cluster from S. coelicolor M145, leaving behind the suicide vector pUC119::neo, a scar of 42 bp attL site and pIJ10500 integrated within SCO4848. Excision of the gene cluster was confirmed by PCR analysis using both genomic and plasmid DNA as templates (Fig. 2). For 9 out of 10 (pSV::attB 6 Up) carrying attB 6 and a region homologous to 59 end of the cluster is introduced into the chromosome by a single crossover. A second plasmid pKC1139::attP 6 Dn is based on pKC1139 carrying attP 6 and a region homologous to 39 end of the cluster. When the incubation temperature is higher than 34uC, pKC1139::attP 6 Dn turns into a non-replicating plasmid and then is integrated into the chromosome by a single crossover. Expression of wBT1 integrase (encoded in the plasmid pIJ10500) leads to excision of the pKC1139 backbone with gene cluster of interest, leaving behind the suicide vector pUC119::neo and 42 bp attL6 site. aac(3)IV: apramycin resistance gene; neo: kanamycin resistance gene; ori: temperature-sensitive origin of replication from pSG5; rep: rep encoding a replication initiator protein from pSG5.
www.nature.com/scientificreports SCIENTIFIC REPORTS | 5 : 8740 | DOI: 10.1038/srep08740 exconjugates tested, the wBT1 integrase-mediated excision of act gene cluster occurred at a frequency of 90%. Furthermore, the presence of attL and attR in the amplified fragments was confirmed by DNA sequencing (Fig. 3). To recover the plasmid containing the entire act gene cluster (pKC1139::act) from Streptomyces, the DNA extract containing pKC1139::act from M145-MCact was used to transform E.coli Top10. Plasmid DNA from four apramycin resistant E. coli colonies was confirmed by BamHI and NotI digestion, respectively. The restriction fragments showed correct band patterns (Fig. S2a).
To verify the cluster is complete, the recombinant plasmid pKC1139::act was introduced into S. coelicolor M1146 (M1146) to obtain M1146-MCact. Unlike M1146 and M1146-pKC1139 (S. coelicolor M1146 containing empty vector pKC1139), M1146-MCact regained the ability to produce the blue pigment actinorhodin (Fig. 4a). These results showed that the cloned act gene cluster was complete and functional.
Deletion of act gene cluster from S. coelicolor M145. To delete the act gene cluster from S. coelicolor M145, a single colony of M145-MCact was randomly chosen for the removal of pKC1139::act. After three rounds of nonselective growth at 28uC and subsequent cultivation at 40uC, approximately 5% of M145-MCact colonies lost pKC1139::act. Strains lacking the act gene cluster (M145-Dact) were first confirmed by PCR (data not shown), and then patched on R5MS solid agar plate for visual comparison of actinorhodin production. Unlike S. coelicolor M145 that could produce both blue pigment actinorhodin and red pigment undecylprodigiosins, M145-Dact could only produce the red pigment undecylprodigiosins (Fig. 4a). This was further validated by no actinorhodin production of M145-Dact in R5MS liquid culture (Fig. 4b).
Cloning and deletion of nap and dap gene cluster from S. roseosporus NRRL 15998. To clone gene cluster of medium and large sizes, we used the same strategy to clone nap and dap gene cluster from S. roseosporus NRRL 15998. Excision of nap gene cluster from S. roseosporus NRRL 15998 occurred in 9 out of 10 exconjugates, and excision of dap gene cluster from S. roseosporus NRRL 15998 occurred in 8 out of 10 exconjugates (Fig. 2). Like pKC1139::act, plasmid containing nap gene cluster (pKC1139::nap) was passed through E.coli Top10 and isolated plasmid DNA was confirmed by BglII and EcoRI digestion, respectively (Fig. S2b). The cloned fragment covers a contiguous DNA region of 45 kb from SSGG02973 to SSGG03009. For pKC1139::dap, the plasmid was isolated directly from Streptomyces and confirmed with restriction digestion (Fig.  S2c). The 157 kb fragment covering SSGG00215-SSGG00287 contains the complete dap gene cluster. Similar to that of act gene cluster, the removal of pKC1139::nap and pKC1139::dap from Sro-MCnap and Sro-MCdap generated strains lacking nap and dap gene clusters (Sro-Dnap and Sro-Ddap).
Improvement of antibiotic titers. The pKC1139 contains a temperature-sensitive origin of replication from pSG5, which is a medium copy plasmid with an approximate 20-50 copy numbers per chromosome 21 . When cultured at 28uC, pKC1139 exists as autonomous plasmid in Streptomyces. In S. coelicolor M145, there is only one copy of act gene cluster in the chromosome. After the wBT1 integrase-mediated excision, the act gene cluster was transferred into pKC1139. An increase in copy number of act gene cluster will improve actinorhodin production. This was confirmed both on R5MS agar plate (Fig. 4a) and in R5MS liquid culture (Fig. 4b). It should be noted that the titer of actinorhodin in M1146-MCact was even higher than that of M145-MCact. Similarly, daptomycin titer www.nature.com/scientificreports could also increase after the excision of dap gene cluster from its chromosome location in S. roseosporus NRRL 15998. Cultures of Sro-MCdap and S. roseosporus NRRL 15998 were subjected to bioassay against S. aureus, the results showed that Sro-MCdap exhibited bigger inhibition zones against S. aureus than S. roseosporus NRRL 15998 at time intervals from 2-5 days (Fig. 5a). This was further verified by comparison of daptomycin from fermentation broth of S. roseosporus NRRL 15998 and Sro-MCdap by highperformance liquid chromatography (HPLC) analysis (Fig. 5b). In addition, we noticed that existence of extra copy numbers of antibiotic gene clusters caused a slowdown in growth of Streptomyces. When cultured on AS-1 agar medium, growth of Sro-MCnap and Sro-MCdap are severely impaired, especially at earlier stages of cultivation (Fig. S3). This phenotype was most likely attributed to the metabolic burden of extra copy numbers of antibiotic gene clusters. This assumption is based on the observation that growth of Sro-Dnap (devoid of nap gene cluster) and Sro-Ddap (devoid of dap gene cluster) are converted back to that of S. roseosporus NRRL 15998 (Fig. S3).
To examine the stability of multiple copy plasmids in Streptomyces, two randomly chosen strains of Sro-MCdap were passed consecutively for five or ten times on AS-1 plates supplemented with or without apramycin. Biological activities of these stains (G 5 and G 10 ) were compared with that of the original Sro-MCdap (G 0 ). All Sro-MCdap strains exhibited similar inhibitory activity against S. aureus (Fig. S4), suggesting that pKC1139-derived large plasmids are stable in the engineered Streptomyces in the presence or absence of selective pressure.

Discussion
We have established an efficient method for genome engineering and direct cloning of gene clusters in Streptomyces. The strategy is based on phage wBT1 attP-attB-int system and provides several advantages over similar methods. First, it can be used for the deletion of large fragment (up to 157 kb) from Streptomyces genome. In the meantime, the large fragment containing gene cluster of interest was cloned into pKC1139 simultaneously. Another advantage of our strategy is that it could clone gene cluster in size up to 157 kb. This is the largest size ever reported in Gram positive bacteria and should be good enough for most antibiotic gene clusters. Last, our strategy can be used to improve the titer of industrial important antibiotics by creating strains with extra copy numbers of antibiotic biosynthetic gene clusters.
The wBT1 attP-attB-int system is helpful for genetic modifications of Streptomyces genome at multiple sites. In addition to the intact attP-attB pair, there are 15 mutated attP-attB pairs which can be recognized by wBT1 integrase 17 . Multiple rounds of large fragment deletion can be achieved with the following modifications.
(1) Relocation of the attB 6 sequence (or any other mutated attB) to the downstream of the 2.0 kb homologous fragment in pSV:: attB 6 Up. This change will allow the excision of the pUC119::neo backbone from Streptomyces genome together with pKC1139.
(2) Construction of an autonomous helper plasmid containing a temperature-sensitive origin of replication from pSG5, the origin of transfer (oriT) from plasmid RK2 and wBT1 integrase gene. This plasmid can ensure the high efficient excision of large fragment from Streptomyces genome and subsequent removal of wBT1 inte- Genome analysis suggested that S. roseosporus NRRL 15998 has potential capacity to produce napsamycins 22,23 . However, the production of napsamycins in S. roseosporus NRRL 15998 has not been reported. With this strategy, we cloned nap gene cluster in pKC1139 to generate pKC1139::nap. It can be manipulated extensively in E. coli. These manipulations include replacement of vector backbone with integrative plasmid and deletion or constitutive expression of regulatory gene by PCR targeting 24 . The modified gene cluster can be transferred into heterologous hosts for expression. It can also be transferred back into the mutant devoid of nap gene cluster after removal of pKC1139::nap. Detection of napsamycin in these strains will shed light on the activation of cryptic gene clusters in Streptomyces.
In some industrial overproducing strains generated by traditional mutagenesis, amplification of biosynthetic gene cluster has been observed [25][26][27] . Based on these observations, controlled amplification of gene cluster was used to increase the productivity of commercially important antibiotics. Integration of an additional copy of gene cluster for nikkomycin and gougerotin biosynthesis led to an increased production of nikkomycin and gougerotin by Streptomyces ansochromogene 28 and Streptomyces graminearus 29 , respectively. The zouA-mediated gene amplification of act gene cluster in S. coelicolor M145 led to a 20-fold increase in actinorhodin production 30 . The zouA encodes a site-specific relaxase similar to TraA protein which catalyzes RecA-independent site-specific recombination. The recombination sites of ZouA are oriT-like RsA and RsB 30 . In this study, we reported the amplification of gene clusters mediated by phage wBT1 integrase and improved antibiotic titers in the engineered Streptomyces strains. We believe that the system described here could be used readily to increase antibiotic titers in many Streptomyces and possible other actinomycetes.

Methods
Bacterial strains, plasmids, primers and growth conditions. Bacterial strains and plasmids used in this study are listed in Table 1, and primers are listed in Table S1. S. coelicolor M145 and S. roseosporus NRRL 15998 were used for cloning of act, nap and dap gene clusters. S. coelicolor M1146 is an engineered derivative of S. coelicolor M145 that lacks gene clusters for actinorhodin (ACT), undecylprodigiosins (RED), cryptic polyketide (CPK) and calcium-dependent antibiotic (CDA) biosynthesis 31 . Staphylococcus aureus was used as an indicator strain for daptomycin bioassay. E. coli Top10 was used as a general host for propagating plasmids. E. coli ET12567 (pUZ8002) was used as a host for transferring DNA from E. coli to Streptomyces by intergeneric conjugation 32 .
For general purpose, S. coelicolor M145 and its derivatives were grown on mannitol soya flour medium (MS) agar or in yeast extract-malt extract (YEME) liquid med-    ium 32 . For actinorhodin production, S. coelicolor M145 and its derivatives were grown on R5MS agar or in R5MS liquid medium 33 . S. roseosporus NRRL 15998 was cultured on AS-1 agar medium or in tryptic soy broth (TSB) liquid medium 34 . All Streptomyces stains were maintained at 28uC unless specified otherwise. General approaches for E. coli or Streptomyces manipulations were performed according to standard protocols 32,35 . When necessary, the final antibiotic concentrations used for selection of E. coli transformants were as follows: ampicillin, 100 mg ml 21 ; apramycin, 100 mg ml 21 ; kanamycin, 100 mg ml 21 ; chloramphenicol, 12.5 mg ml 21 . For selection of Streptomyces transformants, the final antibiotic concentrations were, kanamycin, 50 mg ml 21 in MS for S. coelicolor and 20 mg ml 21 in AS-1 for S. roseosporus; apramycin, 50 mg ml 21 in MS for S. coelicolor and 10 mg ml 21 in AS-1 for S. roseosporus; hygromycin, 50 mg ml 21 in MS or AS-1 for Streptomyces; nalidixic acid, 25 mg ml 21 in MS or AS-1 for Streptomyces.
Construction of plasmids. Of the 15 mutated attP-attB pairs 17 , attP 6 -attB 6 was randomly chosen for this experiment. The sequences of attB 6 and attP 6 were obtained by overlapping PCR. For construction of pSV::attB 6 -act, a 2.0 kb fragment flanking 59 end of the act gene cluster was amplified from genomic DNA of S. coelicolor M145 with primer pair act-Up F/act-Up R. The amplicon was diluted 15100 and used as templates for the second round of PCR with primer pair attB 6 -in F/act-Up R. The product from the second amplification reaction was diluted again and underwent a third run with primer pair attB 6 -out F/act-Up R. The final product was digested with HindIII/BamHI and then inserted into the corresponding sites of pUC119::neo to generate pUC119::neo-attB 6 -act. The origin of transfer (oriT) from plasmid RK2 was amplified from pKC1139 with primer pair oriT F and oriT R, subsequently digested with EcoRI and inserted into the EcoRI site of pUC119::neo-attB 6 -act to generate pSV::attB 6 -act (Fig. S1a). For construction of pKC1139::attP 6 -act, a 2.0 kb fragment flanking 39 end of the act gene cluster was amplified from genomic DNA of S. coelicolor M145 with primer pair act-Dn F/act-Dn R. The amplicon with 15100 dilution served as templates for the second round of PCR with primer pair attP 6 F/act-Dn R. The final product was digested with HindIII/EcoRI and then inserted into the corresponding sites of pKC1139 to generate pKC1139::attP 6 -act (Fig. S1b). For construction of pSV::attB 6 -nap and pSV::attB 6 -dap, a 2.0 kb fragment flanking 59 end of the nap and dap gene cluster was amplified from genomic DNA of S. roseosporus NRRL 15998 with primer pairs nap-Up F/nap-Up R and dap-UpF/dap-Up R, respectively. The product was digested with XbaI/BamHI and used to replace the 2.0 kb fragment upstream of the act gene cluster in pSV::attB 6 -act. For construction of pKC1139::attP 6 -nap and pKC1139::attP 6 -dap, a 2.0 kb fragment flanking 39 end of the nap and dap gene cluster was amplified from genomic DNA of S. roseosporus NRRL 15998 using primer pairs nap-Dn F/nap-Dn R and dap-Dn F/dap-Dn R, respectively. The product was digested with BamHI/EcoRI and used to replace the 2.0 kb fragment downstream of the act gene cluster in pKC1139::attP 6 -act. To ensure the authenticity of DNA sequences, all PCR products were verified by sequencing.
Construction of double-cointegrate strains. To insert attB 6 and attP 6 at sites flanking the act gene cluster of S. coelicolor, pSV::attB 6 -act and pKC1139::attP 6 -act were conjugated into S. coelicolor M145. The pSV::attB 6 -act is unable to replicate alone in Streptomyces and selection with kanamycin allows to select exconjugants in which pSV::attB6-act is inserted into the S. coelicolor genome. The pKC1139::attP 6act is a derivative of the E. coli-Streptomyces shuttle vector pKC1139 that contains a Streptomyces temperature-sensitive origin of replication from pSG5 15 . When the incubation temperature is higher than 34uC, pKC1139::attP 6 -act turns into nonreplicating plasmid and attP 6 was then inserted into S. coelicolor genome with selection of apramycin to obtain double-cointegrate strain Sco-actB 6 P 6 . Similar strategy was used for the construction of double-cointegrate strains Sro-napB 6 P 6 and Sro-dapB 6 P 6 .
Excision of targeted regions. The integrative plasmid pIJ10500 36 is a derivative of pMS82 which contains the phage wBT1 integrase gene and integrates intragenically into SCO4848 encoding a putative integral membrane protein 16 . It was conjugated into a randomly selected strain Sco-actB 6 P 6 , Sro-napB 6 P 6 and Sro-dapB 6 P 6 , respectively. The exconjugants were initially selected with hygromycin and ten randomly chosen exconjugants were passed twice on MS or AS-1 plates supplemented with kanamycin and apramycin, and subject to genomic and plasmid extraction. Excision of targeted region from Streptomyce genome was analyzed by PCR amplifications using genomic DNA templates and primer pairs B 6 -VF/actDn-VR, B 6 -VF/napDn-VR and B 6 -VF/dapDn-VR. In the meantime, PCR amplifications were performed with plasmid DNA template by using primer pairs P 6 -VF/actUp-VR, B 6 -VF/napUp-VR and B 6 -VF/dapUp-VR. Strains with excision of the targeted regions were designated as M145-MCact, Sro-MCnap and Sro-MCdap, respectively.
Deletion of the antibiotic gene clusters from Streptomyces. A single colony of M145-MCact was randomly chosen and passed three times on nonselective MS plates at 28 uC. Spores were harvested, serially diluted, and then spread on MS agar. After growing for 4 days at 40uC, colonies were replicated on MS agar plates containing kanamycin or apramycin. Strains lacking the act gene cluster (M145-Dact) are apramycin sensitive (Apr s ) and kanamycin resistant (Kan r ). Apr s and Kan r strains were further verified by PCR. In a similar way, strains lacking the nap and dap gene clusters (Sro-Dnap and Sro-Ddap) were obtained by the removal of pKC1139::nap and pKC1139::dap from Sro-MCnap and Sro-MCdap.
Actinorhodin quantification. To quantitate actinorhodin production, S. coelicolor M145 and its derivatives were grown in 50 ml of R5MS at 28uC. 1 ml culture was harvested in a time-course and treated with KOH (1 N final concentration), and titer was calculated by measuring the absorbance at 640 nm 37 .
Production and analysis of daptomycin. Small-scale fermentation of daptomycin was carried out by following the procedures described previously 38 For daptomycin analysis, culture broths were centrifuged at 13,000 3 g for 10 min to remove the mycelia. The supernatants were filtered through a Millipore membrane (pore diameter, 0.22 mm) and 50 ml of sample was used for HPLC analysis. Separation of daptomycin was achieved with an Agilent 1100 HPLC system and a ZORBAX SB-Aq column (5 mm pore size, 4.6 by 250 mm). HPLC conditions were described as follows: gradient elution with buffer A (0.01% [vol/vol] trifluoroacetic acid in acetonitrile) and buffer B (0.01% [vol/vol] trifluoroacetic acid in ddH 2 O), flow rate at 1.0 ml/min, ultraviolet detection at wavelength of 224 nm. The elution profile was a linear gradient of 10%-100% buffer A over 22 min, a hold at 100% buffer A over 3 min, a linear gradient of 100%-10% buffer A over 2 min and a final hold at 10% buffer A over 3 min.
Bioassay against S. aureus was performed as previously described with modifications 38 . In brief, S. roseosporus and its derivatives were patched on AS-1 agar. After incubation for 2-5 days at 28uC, agar plugs were prepared from the patches, placed on the surface of an empty Petri dish, and overlaid with culture of indicator strain in soft nutrient agar containing 5 mM CaCl 2 . The zone of inhibition was assessed after overnight incubation at 37uC.