Introduction

The so-called rare or uncommon actinomycetes include filamentous actinomycetes other than Streptomyces that are difficult to isolate, cultivate and genetically manipulate [1]. Some of these difficult-to-handle microorganisms produce medically valuable antibiotics. However, the study and cost-effective exploitation of uncommon actinomycetes, including both strain and product improvement, have been hampered by a lack of genetic tools. The presence of potent restriction-modification systems and the absence of efficient transformation protocols have particularly hindered the development of genetic systems in many actinomycetes [2].

While procedures for protoplast transformation were successfully developed for Streptomyces spp. [3], they are not generally applicable to other actinomycetes [4–7]. Transformation by electroporation has been reported, but it is of low efficiency and of limited application [2]. We recently developed a system for protoplast preparation, fusion and regeneration for uncommon actinomycetes producing metabolites of industrial and medical interest [8, 9]. There has also been considerable interest in the use of intergeneric conjugation as a means of transferring single-stranded DNA, which evades restriction barriers, into actinomycetes, even if procedures have to be customised for individual species before achieving acceptable conjugation efficiencies [4, 10]. Here we describe the development of methods for the manipulation of Nonomuraea sp. ATCC 39727. This actinomycete belongs to the Streptosporangiaceae family and produces the glycopeptide antibiotic A40926, the precursor of the semi-synthetic derivative dalbavancin, a second-generation glycopeptide currently in clinical development [11, 12]. Methods for the genetic manipulation of the A40926-producing strain will allow knowledge-based improvements in antibiotic productivity and the application of combinatorial biosynthesis to develop novel, potentially improved A40926 derivatives.

Materials and methods

Bacterial strains, vectors and recombinant DNA methods

Nonomuraea sp. ATCC 39727 was used as the recipient strain throughout this study. E. coli strain DH5α (Promega) was used as general host for cloning. For transformation and conjugation experiments, plasmids were prepared and maintained as follows. The site-specific integration vector pSET152 (5.7 kb) was extracted from E. coli strains S17-1 [13] and ET12567 [14] using the Qiagen Midi DNA extraction kit (Qiagen). The integrative plasmid pSET152 contains ΦC31 int, attP and oriT of RK2, as well as an apramycin resistance gene for selection in actinomycetes and E. coli [15]. pSET152 is not able to replicate autonomously in actinomycetes, and can be maintained in recipient strains only by integration at a chromosomal attB site. The Supercosl derivative cosmid A40ΔY (30 kb) [16] contains the T3, T7 and SV40 promoters, the pUC origin of replication, and ampicillin and neomycin resistance genes for selection in E. coli and some actinomycetes, respectively. Since cosmid A40ΔY lacks an origin of replication for actinomycetes, its maintenance in Nonomuraea sp. ATCC 39727 requires homologous recombination with the chromosome mediated through the inserted 22 kb segment of Nonomuraea sp. ATCC 39727 DNA. Cosmid extraction was performed by alkaline lysis following the method reported by Kieser [3]. The methylation-deficient E. coli strain ET12567 (dam-13::Tn9, dcm-6, hsdM, hsdS) containing the non-transmissible helper plasmid pUZ8002 was used as the donor strain in biparental intergeneric conjugations [17].

Media and culture conditions

Nonomuraea sp. ATCC 39727 was maintained as a lyophilised master cell bank (MCB). A working cell bank (WCB) was prepared from a first generation slant originating from the MCB as previously described [8, 18]. Mycelium for protoplast preparation was grown in 100 ml VM liquid medium [8] in 500 ml Erlenmeyer baffled flasks incubated at 28°C and 200 rpm. Growth was measured as dry weight as previously described [18]. Conditions for growth and protoplast regeneration of Nonomuraea sp. ATCC 39727 were essentially as described in Marcone et al. [8]. In brief, agar medium for growth of Nonomuraea sp. ATCC 39727 was V0.1 medium, and the hypertonic medium used for protoplast regeneration was M3 overlaid with VMS0.1 [8]. VSP [8] was used to grow Nonomuraea sp. ATCC 39727 to prepare biomass for intergeneric conjugation. Agar media assessed for conjugation were V0.1, V0.1 supplemented with 10 mM MgCl2, and M3. For the extraction of chromosomal DNA from the transformants/exconjugants of Nonomuraea sp. ATCC 39727, antibiotic-resistant clones were inoculated into 10 ml VSP medium containing 100 μg/ml apramycin. Cultures were grown for three days at 30°C, then a further 10 ml of the same medium were added to each flask. Growth was continued for an additional 24 h. Samples were collected by centrifugation and DNA extraction was performed according to the Kirby procedure [3]. E. coli strains were grown at 37°C for 12 h on a rotary shaker (180 rpm) in Luria Bertani medium (LB, Difco) supplemented with 50 μg/ml apramycin, 25 μg/ml chloramphenicol, 100 μg/ml ampicillin and 50 μg/ml kanamycin as required.

Preparation, regeneration and protoplast transformation

Protoplasts of Nonomuraea sp. ATCC 39727 were prepared and regenerated as already described [8]. For transformation, protoplasts were concentrated to 2–4 × 109 in 100 μl of P medium and different amounts of plasmid or cosmid DNA (in TE buffer) added in a maximum volume of 40 μl. 1 ml of 25% polyethylene glycol (PEG) 1000 or LipofectamineTM 2000 (Invitrogen) in P medium was added to the protoplast/DNA mixture, followed by incubation for a maximum of two min at room temperature. The transformation mix was centrifuged in a table-top centrifuge for 1 min and immediately suspended in 700 μl of P medium. 0.2 ml of the transformation mixture was plated on each M3 plate and overlaid with 4 ml of melted pre-cooled VMS0.1 medium. Protoplasts were allowed to regenerate at 30°C for one to three days and were then overlaid with 4–8 ml of VMS0.1 medium supplemented with apramycin at 50 μg/ml. Plates were incubated for an additional 10–15 days until resistant colonies appeared. Resistant colonies were then replicated with a toothpick onto fresh V0.1 medium plates supplemented with 100 μg/ml apramycin. After 5–6 days of growth, colonies were inoculated in liquid medium for DNA extraction as reported above.

Conjugation

Intergeneric conjugation was achieved as suggested by Kieser et al. [3] with modification. In brief, a culture of the donor E. coli ET12567/pUZ8002 containing the selected plasmid was grown in 10 ml LB supplemented with 50 μg/ml apramycin, 25 μg/ml chloramphenicol and 50 μg/ml kanamycin to an OD600 of 0.4. Cells were collected by centrifugation, washed twice with an equal volume of LB, and resuspended in 1 ml of LB. Nonomuraea sp. ATCC 39727 mycelium was prepared as follows. The strain was grown in 20 ml VSP medium in a 100 ml baffled flask containing 2 ml sterile glass beads (diameter 4 mm) and incubated at 28°C on a rotary shaker at 200 rpm to exponential phase (about 24 h). Mycelium was collected and washed twice with ice-cold glycerol 20%. After centrifugation, the pellet was suspended in half the initial volume of glycerol 20% and stored at −80°C. For conjugation, 0.5 ml E. coli donor cells (ca. 1 × 107 cells/ml) were added to the mycelium (0.5 ml) and the mixtures were spread on V0.1 plates containing 10 mM MgCl2. The conjugation plates were incubated for 20–22 h at 30°C, and were then overlaid with 1 ml water containing 1 mg nalidixic acid and 1.25 mg apramycin. The plates were incubated at 30°C for 10–15 days. To confirm the chromosomal integration of the pSET152 or of the A40ΔY cassette, the exconjugants were analyzed by colony-PCR and sequencing of PCR products.

Colony PCR for Nonomuraea sp. ATCC 39727

Single colonies (streaked from ex-conjugants) were patched onto V0.1 medium. Plates were incubated at 30°C for 3–5 days. Mycelium was scraped from the plates using a sterile toothpick and was introduced into 50 μl 100% DMSO (Sigma-Aldrich) in a 1.5 ml tube containing two sterile glass beads. The tube was shaken vigorously for 1–2 h and then centrifuged briefly to pellet cell debris. 2.5 μl of the supernatant was used instead of the 2.5 μl of DMSO in a standard colony PCR. Control DNA samples were diluted 1 μl into 1.5 μl DMSO and were used in the same way. One initial step of 10 min at 95°C was included in the PCR program to ensure the complete cell lysis. The PCR was performed for 30 cycles as follows: 95°C for 45 s, 56°C for 30 s and 72°C for 1 min.

pSET152F and pSET152R were used as primers to amplify an internal fragment of pSET152 [19]. Primers C31F (5′-CTGCTGCACCAGCCACTC-3′) and C31R (5′-GATGGGCGAGGTCGACTAC), designed from the sequence surrounding the ΦC31 attB site of Nonomuraea sp. (GenBank: AJ564722.1), and LF45F and LF45R [19], were used to amplify across the predicted junctions formed by the integration of pSET152 at the Nonomuraea sp. ΦC31 attB site, thus confirming the integration of the plasmid in the Nonomuraea sp. chromosome.

Results and discussion

Protoplast transformation with pSET152 or cosmid A40ΔY

Protoplast transformation was initially performed with the integrative plasmid pSET152 (5.7 kb) that contains the bacteriophage ΦC31 att/int site-specific integration system. The attB integration site of Nonomuraea sp. ATCC 39727 has been previously sequenced and found to be homologous to the corresponding sequences found in Streptomyces [20], as well as to the recently reported attB site in Actinoplanes teichomyceticus [10]. Nonomuraea sp. ATCC 39727 was reported to be resistant to antibiotics such as chloramphenicol, erythromycin and hygromycin, which are commonly used for selection [20]. In contrast, the minimal inhibitory concentration of apramycin for Nonomuraea sp. ATCC 39727 on agar medium is less than 1 μg/ml, permitting the use of the apramycin resistance gene of pSET152 as selectable marker.

Preliminary experiments indicated that transformation was much more efficient with plasmid DNA extracted from the non-methylating E. coli ET12567 than from the methylating E. coli S17-1 (data not shown), suggesting that Nonomuraea sp. ATCC 39727, like many other actinomycetes, contains a methylated-DNA-dependent restriction system [10, 15]. Protoplasts were prepared from mycelium grown to both the exponential (ca. 20 h) and stationary phases of growth (ca. 48 h): nearly 50% of the protoplasts regenerated from mycelium harvested during active growth while 30% regenerated from stationary phase cells. PEG 1000 or LipofectamineTM 2000 was used to facilitate transformation (the latter is often employed to transfect eukaryote cells [21]). The results are shown in Table 1. To confirm that the apramycin resistant colonies were transformants rather than spontaneous apramycin resistant mutants, PCR amplification of the aac(3)IV gene of pSET152 was performed on genomic DNA from 10 randomly selected apramycin resistant clones. All of the selected clones yielded a DNA fragment corresponding in size to the fragment amplified from the control pSET152 (data not shown). Nonomuraea sp. ATCC 39727 genomic DNA was used as the negative control. Since pSET152 cannot replicate autonomously in actinomycetes, it is assumed that the plasmid had integrated stably at the ΦC31 attB site (see Conjugal transfer of pSET152 below).

Table 1 pSET152 transformation of Nonomuraea sp. ATCC 39727
Full size table

As expected from the regeneration efficiencies, protoplasts made from the exponential phase culture gave markedly higher transformation frequencies than those prepared from stationary phase mycelium. While there was no consistent difference in transformation efficiency between PEG 1000 and LipofectamineTM 2000, transformation in the presence of LipofectamineTM 2000 gave optimal results when using 2 μg of plasmid DNA (3.7 × 103 transformants per μg of plasmid DNA; Table 1). This presumably reflects that LipofectamineTM 2000 better complexes higher concentrations of nucleic acids, allowing the transformation with higher DNA concentrations in comparison to PEG [21]. The highest efficiencies obtained here are markedly lower than those reported for Streptomyces spp. (106 transformants per μg of plasmid DNA) [3].

Protoplast transformation was then performed with the Supercos-1 cosmid derivative A40ΔY designed to inactivate the vanY gene putatively involved in self-resistance to A40926 in Nonomuraea sp. ATCC 39727 [16]. A40ΔY contains a 22 kb segment of the Nonomuraea sp. ATCC 39727 chromosome and does not replicate in Nonomuraea. Thus, stable transformants can only be obtained by homologous recombination.

Cosmid A40ΔY was extracted from the non-methylating E. coli ET12567 and used to transform protoplasts following the protocol described above for transformation with pSET152. Different protoplast and DNA concentrations were used. However, no transformants were obtained. This likely reflects the larger size of the plasmid used and/or to the different method of integration into the recipient genome. Denaturation by alkali treatment of the cosmid just before transformation, as described in [3], also failed to yield transformants.

Conjugal transfer of pSET152

In order to assess whether we could transfer large segments of DNA into Nonomuraea sp. ATCC 39727, we first developed a conjugation protocol using the small integrative plasmid pSET152. Initially, intergeneric conjugation between Nonomuraea sp. ATCC 39727 and E. coli ET12567/pUZ8002 containing pSET152 was attempted by adapting the protocol used with Streptomyces spores [3]; since conditions for sporulation of Nonomuraea sp. ATCC 39727 have not been determined, mycelium was used as the recipient.

As reported for protoplast production [8], it was first necessary to determine conditions that gave dispersed growth of Nonomuraea sp. ATCC 39727 in liquid culture. SFM, ISP4 and RARE3 [3, 20] liquid media all yielded compact pellets of mycelium not suitable for conjugation. Only VSP medium (with glass beads added to the flasks) gave dispersed growth. Furthermore, we observed that freezing of Nonomuraea sp. ATCC 39727 mycelium before conjugation enhanced dispersion. The agar medium used to select exconjugants can greatly influence the efficiency of conjugation in actinomycetes [10, 22]. Therefore, six representative media were tested. We selected medium V0.1 (routinely used to grow Nonomuraea sp. ATCC 39727), M3 (previouly optimized for protoplast regeneration [8]), BTT (already reported in another conjugation protocol [20] for Nonomuraea sp. ATCC 39727), and MS (suitable for streptomycetes and more recently for the non-Streptomyces actinomycete Actinoplanes teichomyceticus [10]). To increase the conjugation efficiency of Streptomyces strains, 10 mM MgCl2 is generally added to agar media [3], although the function of Mg2+ has not been investigated in detail. Therefore, 10 mM MgCl2 was added to medium V0.1 and MS (M3 already contained 10 mM MgCl2). The results reported in Table 2 show that the frequency of pSET152 conjugation (calculated as the number of apramycin resistant colonies per Nonomuraea sp. ATCC 39727 CFU determined from the biomass used for conjugation) varies widely depending on the agar medium used. With MS (the most commonly used media for streptomycetes [3, 22] and for Actinoplanes teichomyceticus [10]), we did not obtain any exconjugants, whereas in media specifically developed for Nonomuraea sp. ATCC 39727 [8], we had the highest number of resistant clones. This suggests that the species-specific fine-tuning of medium composition is crucial in conjugation as well as in protoplast regeneration [8]. The addition of MgCl2 plays a positive role when added to V0.1, increasing conjugation efficiency several fold and consistent with previous observations for Actinoplanes teichomyceticus [10]. The best medium (V0.1 supplemented with 10 mM MgCl2) gave 4,040 colonies resistant to 50 μg/ml of apramycin from a conjugation mixture of 500 μl of Nonomuraea sp. ATCC 39727 (1 × 107 CFU/ml) and 500 μl of E. coli ET12567/pUZ8002 (OD600 of ca. 0.4, 1 × 107 cells/ml). One hundred randomly chosen colonies resistant to apramycin were analysed by PCR. Figure 1 shows amplification of a DNA fragment corresponding to the aac(3)IV gene in four representative clones, indicating that they are true exconjugants and not spontaneous resistant mutants. This was confirmed by the analysis of the integration site of pSET152 (Fig. 1).

Table 2 Effect of media on intergeneric conjugation with pSET152
Full size table
Fig. 1
figure 1

PCR analysis of genomic DNA from 4 putative Nonomuraea sp. pSET152 exconjugants (1–4 and 8–15) compared to Nonomuraea sp. ATCC 39727 wild type genomic DNA (7) and pSET152 DNA (5–6). Lanes 1–5 show the PCR products obtained with primers pSET152F and pSET152R, lane 6 the PCR product of primers LF45F and LF45R, lanes 7–11 the PCR products of primers LF45F and C31R, and lanes 12–15 the PCR products of primers C31F and LF45R. The marker (Molecular Weight, MW) was a 1 kb ladder (NEB). The following expected band sizes were observed (based on the sequence information): 319 bp using pSET152F and pSET152R, 74 bp using LF45F and C31R, and 255 bp using C31F and LF45R, confirming integration of pSET152 at the ΦC31 attB

Full size image

The frequency of 10−3 exconjugants per recipient Nonomuraea sp. ATCC 39727 CFU is comparable with data reported for the same vector in other actinomycetes, such as Actinoplanes teichomyceticus (~10−3) [10] and Streptomyces pristinaespiralis (~10−3) [22]. Use of the ΦC31 att/int system in Streptomyces lividans is reported to give a higher frequency (8 × 10−2) [3, 10].

Conjugal transfer of cosmid A40ΔY

The protocol developed for the conjugal transfer of pSET152 was used to introduce cosmid A40ΔY from the non-methylating E. coli ET12567/pUZ8002 into Nonomuraea sp. ATCC 39727. In these conditions (medium V0.1 supplemented with 10 mM MgCl2; 500 μl of Nonomuraea sp. (1 × 107 CFU/ml) and 500 μl of E. coli ET12567/pUZ8002 (1 × 107 CFU/ml)), 55 colonies resistant to 50 μg/ml of apramycin were obtained, i.e. 1.1 × 10−5 exconjugants per Nonomuraea sp. ATCC 39727 CFU. Table 2 shows that there is a reduction in conjugation efficiency of about one hundred fold in comparison to pSET152, likely due to the larger size of the transferred plasmid and/or the method of chromosomal integration (i.e. site specific integration vs homologous recombination).

From the PCR/agarose gel electrophoresis analysis of the 55 apramycin resistant colonies (as described in "Materials and methods"), 37 showed a PCR amplification profile consistent with integration of cosmid A40ΔY into the Nonomuraea sp. ATCC 39727 genome by a single homologous recombination event, while 18 colonies showed a PCR amplification profile consistent with a double cross-over event that replaced vanY with the apramycin resistance cassette (for details see [23]). Figure 2 shows the pattern of bands resulting from colony-PCR performed from a sample of 18 exconjugants. Cosmid A40ΔY and genomic DNA from Nonomuraea sp. ATCC 39727 were used as positive and negative controls, respectively.

Fig. 2
figure 2

Amplification of the inactivation cassette (containing aac(3)IV flanked by two oriT and FLP sites) in apramycin resistant clones from conjugation experiments with cosmid A40ΔY. Cosmid A40ΔY DNA was used as positive control (C), while genomic DNA from Nonomuraea sp. ATCC 39727 wild-type was used as negative control (G1 and G2). 1 kb NEB: molecular weight markers. Two bands (1,624 and 800 bp), indicating that a single homologous recombination event occurred following conjugation, are present in eight clones tested, while 10 clones show a single band (1,624 bp) corresponding to two homologous recombination events (i.e. a double cross-over)

Full size image

Minimal inhibitory concentration (MICs) of A40926 were determined for five of the 18 positive clones: all gave a MIC value of 2 μg/ml for A40926 rather than the 4 μg/ml of the wild-type strain, thus confirming inactivation of vanY and its involvement in self-resistance to A40926 in Nonomuraea sp. [16].

Conclusions

Two gene transfer systems were developed and optimized for Nonomuraea sp. ATCC 39727. The first one is based on protoplast production, transformation and regeneration, and requires de-methylated DNA. The second is based on E. coli ET12567/pUZ8002-Nonomuraea sp. ATCC 39727 intergeneric conjugation and was used successfully with both a 5.7 kb plasmid DNA and a 30 kb cosmid.

With the site-specific integrative plasmid pSET152, protoplasts were transformed with an efficiency of 1.5 × 103 per μg of plasmid DNA and 1.5 × 10−5 transformants per regenerating protoplast, whereas conjugation efficiency was estimated to be 10−3 per Nonomuraea sp. ATCC 39727 CFU.

With cosmid A40ΔY, which requires homologous recombination, the efficiency of conjugation was 1.1 × 10−5 per Nonomuraea sp. ATCC 39727 CFU, and was sufficient to isolate both single and double homologous recombination events.

The gene transfer systems developed here for Nonomuraea sp. ATCC 39727, combined with the recently reported use of Nonomuraea sp. as a host for the heterologous expression of an antibiotic biosynthetic gene cluster [19], further demonstrate the ability to genetically manipulate this previously relatively intractable actinomycete. The methods described here can now be used to better understand A40926 biosynthesis and self-resistance, and subsequently to manipulate the producing organism to enhance productivity and potentially produce A40926 derivatives with improved characteristics.