Sonication-Dependent Electroporation of the Erythromycin-Producing Bacterium Saccharopolyspora erythraea

Abstract

We report the development of an electrotransformation method applicable to all strains of Saccharopolyspora erythraea examined to date. Vegetatively grown mycelia were rendered electrocompetent by subjecting mycelial suspensions to ultrasound pulses. The protocol provides an alternative route for the introduction of DNA into filamentous microorganisms otherwise recalcitrant to transformation techniques.

Saccharopolyspora erythraea, a gram-positive, actinomycete that produces the clinically important macrolide antibiotic erythromycin A, has received considerable attention as a model system for the study of polyketide biosynthesis (2, 3). Molecular genetic manipulations of the S. erythraea polyketide synthase genes have improved our understanding of this complex enzyme system and led to the production of a number of novel macrolide compounds (4, 6). The industrial use of S. erythraea for the fermentative production of erythromycin A has increased in recent years due to the growing demand for the semisynthetic production of a variety of second-generation erythromycin species, including clarithromycin, azithromycin, roxithromycin, and dirithromycin.

Protoplast transformation techniques have been successfully used for the molecular genetic manipulation of a variety of actinomycetes, including wild-type strains of S. erythraea (12). However, during the course of our work with S. erythraea, we have isolated a wide range of mutant strains affecting central and secondary metabolism for which protoplast transformation methods have proven ineffective. Although electroporation has found widespread application for the introduction of DNA into a range of cell types, there have been few reports describing the application of this technology to industrially important filamentous microorganisms (7, 9). We report here an electroporation protocol which yields 10² to 10⁴ transformants/μg of DNA. In addition, we have also used this method for the introduction of DNA into other actinomycetes, including Streptomyces lividans, Streptomyces rimosus, and Saccharopolyspora hirsuta.

Bacterial strains, plasmids, and bacteriophage.

Bacterial strains, plasmids, and bacteriophage are listed in Table 1. Plasmid pCD1 is a shuttle plasmid which contains a colE1 replication origin, a derivative of the pJV1 replicon from Streptomyces phaeochromogenes (1), and markers conferring resistance to ampicillin and thiostrepton (tsr). pMBE2 is a derivative of pCD1 in which the tsr gene was replaced by the hyg gene, which confers hygromycin resistance. The S. erythraea virulent bacteriophage φABT1 was purified from liquid lysate cultures of S. erythraea with CsCl gradients, and the DNA was prepared as previously described (5, 10).

TABLE 1.

Bacterial strains, plasmids, and bacteriophage used in this study

Bacterial strain, plasmid, or bacteriophage	Relevant descriptiona	Source or reference
Strains
Saccharopolyspora erythraea CA340	Erythromycin overproducer	Abbott Laboratories Collection
Saccharopolyspora hirsuta ATCC 27875	Moderate thermophile	American Type Culture Collection
Streptomyces lividans TK24	Contains no native plasmids	John Innes Institute, Norwich, United Kingdom
Streptomyces rimosus ATCC 10970	Oxytetracycline producer	American Type Culture Collection
Escherichia coli DH5α	hsdR17 (rK− mK+)	Life Technologies, Inc.
Escherichia coli GM2929	hsdR2 (rK− mK+) dam-13::Tn9 dcm-6	CGSCb
Plasmids
pCD1	6.5-kb E. coli-Streptomyces plasmid (pJV1 replicon) conferring Thir	C. V. Dery
pMBE2	7.2-kb derivative of pCD1 with hyg substituted for tsr; confers Hygr	M. Babcock
pWHM4	7.2-kb E. coli-Streptomyces bifunctional plasmid conferring Thir	11
pIJ702	5.6-kb high-copy-number plasmid conferring Thir; pIJ101 derivative	John Innes Institute, Norwich, United Kingdom
Bacteriophage φABT1	38-kb virulent bacteriophage of S. erythraea	Abbott Laboratories Collection

^aAbbreviations: Thi^r, thiostrepton resistance; Hyg^r, hygromycin resistance. 

^bE. coli Genetic Stock Center, Yale University, New Haven, Conn. 

Media and buffers.

ABB1 base medium contained the following components (in grams per liter): Bacto Soytone, 10; yeast extract, 9; soluble starch, 10; and cerelose, 15. Filter-sterilized 3-N-morpholinopropanesulfonic acid (MOPS) buffer, pH 7.0, was added to ABB1 base medium after autoclaving to a final concentration of 50 mM. ABB1 soft agarose (0.6% [wt/vol] type VII low-gelling-temperature agarose) was used for plate overlays. ABB13 medium contained the following components (in grams per liter): Bacto Soytone, 5; soluble starch, 5; CaCO₃, 3; MOPS buffer, 2.1, and Bacto Agar, 20. After being autoclaved, 1.0 ml of filter-sterilized 1.0% (wt/vol) stock solution of thiamine HCl and 1.0 ml of 1.2% (wt/vol) FeSO₄ · 7H₂O were added. Phage buffer contained 10 mM TES [N-tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid] buffer (pH 7.5), 10 mM MgCl₂, and 10 mM CaCl₂. Thiostrepton and hygromycin were used at final concentrations of 10 and 80 μg/ml, respectively. S. erythraea CA340 (an industrially improved strain for erythromycin production), S. hirsuta ATCC 27875, and Streptomyces lividans TK24 were cultivated at 33°C in ABB1. Streptomyces rimosus ATCC 10970 was cultivated at 28°C in CRM medium (9). Electroporation wash buffer contained 0.3 M mannitol and 3 mM HEPES, pH 7.0. Electroporation buffer consisted of electroporation wash buffer containing 25% (wt/vol) polyethylene glycol 3350 (Sigma Chemical Co., St. Louis, Mo.). Digoxigenin-labeled DNA probes were prepared as recommended by the vendor (Boehringer Mannheim, Indianapolis, Ind.).

Preparation of electrocompetent S. erythraea cultures.

Electrocompetent S. erythraea CA340 mycelia were prepared by inoculating 10 ml of ABB1 broth containing 0.5% (wt/vol) glycine with 0.1 ml of a frozen glycerol stock culture. The cultures were incubated on a reciprocal shaker at 250 rpm. After achieving an optical density at 600 nm of 2 to 4, the mycelia were sonicated with a Heat Systems-Ultrasonics, Inc., model W-375 sonicator equipped with a tapered microtip at an output of 4.5 and with a 50% duty cycle for 60 pulses. The sonicated mycelia were centrifuged at 12,000 × g for 10 min at 4°C, washed twice in 20 ml of chilled electroporation wash buffer, finally resuspended in 0.6 to 0.8 ml of chilled electroporation buffer per 10 ml of starting culture, and stored on ice until use.

Electroporation assays.

To avoid potential problems with host restriction of Escherichia coli-propagated shuttle vectors, an electrotransfection assay with DNA from the S. erythraea virulent bacteriophage φABT1 was used to define electroporation conditions. Aliquots (0.2 ml) of electrocompetent cells were mixed thoroughly with 100 to 200 ng of bacteriophage φABT1 DNA. The mixtures were transferred to chilled 0.2-cm-gap electrocuvettes (Bio-Rad Laboratories, Hercules, Calif.) and electroporated with a Bio-Rad Gene Pulser and Pulse Controller unit (25-μF capacitor, 600-Ω resistance). Field strengths were varied from 7.5 to 10.0 kV/cm. Following delivery of the electric pulse, the cuvette was removed from the electroporation chamber and incubated at room temperature (28°C) for 40 to 60 min. A 0.8-ml aliquot of ABB1 broth was then added to the cuvette, and the entire mixture was transferred to 3.0 ml of ABB1 soft agarose containing nonelectroporated S. erythraea indicator cells and plated on ABB13 plates. The electrotransformation assay was performed as indicated above except that 100 to 200 ng of pCD1 or pMBE2 DNA was used in lieu of φABT1 DNA and no indicator cells were added. After 22 to 24 h of incubation, antibiotic was added in an additional 3.0 ml of ABB1 soft agarose overlay.

Optimization of electrocompetence.

Initial attempts to electroporate a variety of mutant strains of S. erythraea from the Abbott culture collection were unsuccessful. This group of strains included the industrially improved mutant CA340. In an effort to obtain more homogeneous dispersed culture preparations of strain CA340, the vegetative mycelia were subjected to a brief ultrasonic treatment. This treatment generated well-dispersed uniform hyphal fragments (Fig. 1). Transfectants and transformants were readily obtained with φABT1 DNA and pCD1 DNA. Subsequent experiments indicated that the electrocompetence was strictly sonication dependent. No transfectants or transformants were obtained if either the DNA or the electric pulse was omitted. The number of sonication pulses was optimized, while the sonicator energy output and duty cycle were held constant (Fig. 2). Under these conditions, 60 pulses yielded the maximum number of electroporants with both φABT1 (1.2 × 10³ PFU per μg of DNA) and pCD1 (1.0 × 10⁴ thiostrepton-resistant transformants per μg of DNA). The plateau in electrotransformants is most likely related to the heterogeneity of sonicated hyphal fragment populations. The positive effect of sonication on electrocompetency was eliminated when the sonicated hyphal fragments were returned to culture tubes and incubated with shaking prior to electroporation. Within 60 min of outgrowth prior to electric pulsing, no transformants were recovered. This result suggests that the physical alteration responsible for electrocompetence was eliminated or repaired by the cells during this period. When S. erythraea was grown in the presence of 0.5% (wt/vol) glycine, the transformation efficiency was increased approximately twofold (data not shown). Higher concentrations of glycine inhibited growth and led to deleterious mycelial pellet formation and longer growth times. Pretreatment of S. erythraea mycelia with lysozyme did not enhance electrocompetence.

FIG. 1.

Effect of sonication on S. erythraea mycelia. Untreated vegetatively grown mycelia (left) and mycelia subjected to 60 sonication pulses (right).

FIG. 2.

Effect of the number of sonication pulses on electrotransformation efficiency. Electroporations were carried out with 125 ng of pCD1 (•) or 145 ng of φABT1 DNA (○). The electrical parameters were 1.5 kV, 25 μF, and 600 Ω. The data are the means of results of two replicate electroporation experiments.

The highest electrotransformation efficiency for pCD1 (2.3 × 10⁴ thiostrepton-resistant transformants per μg of DNA) was obtained with a field strength of 8.75 kV/cm (Fig. 3). At this voltage with a capacitance setting of 25 μF, there was a concomitant decrease in cell viability to 45% survival. The optimal field strength for φABT1 DNA was between 7.5 and 10.0 kV/cm (data not shown). Under these conditions, the electrotransfection efficiency was 9.6 × 10² PFU per μg of DNA. To examine the effect of the time constant on electroporation efficiency, the resistance was varied between 200 and 1,000 Ω. This resulted in time constants ranging from 4.1 to 16.5 ms. The optimal time constant was about 11.0 ms at a resistance setting of 600 Ω on the Bio-Rad Pulse Controller unit.

FIG. 3.

Effect of the initial voltage of the applied electric pulse (field strength) on transformation efficiency (•) and the survival (▵) of electroporated mycelia at 25 μF and 600 Ω. Electrotransformations were carried out with 125 ng of pCD1. The data are the means of results of two replicate electroporation experiments.

Plasmid pCD1 stability and effects of DNA modification.

The host restriction modification system of S. erythraea is at present poorly understood. The pJV1 replicon contained within pCD1 is poorly maintained by S. erythraea and is lost in the absence of selective pressure (12). Consequently, it was not possible to obtain significant amounts of pCD1 DNA from S. erythraea cultures. Nonetheless, the transient maintenance of this replicon permitted the recovery of primary thiostrepton-resistant colonies prior to the eventual loss of the plasmid. Loss of the plasmid occurred after subsequent passage of the culture even in the presence of thiostrepton selection. Despite the poor stability of the S. erythraea pCD1 transformants, the pCD1-based electrotransformation assay permitted us to optimize electroporation conditions for plasmid DNA uptake.

We examined the effects of pCD1 DNA modification on electrotransformation efficiency by altering the methylation pattern of this substrate. pCD1 DNA was prepared from E. coli DH5α and E. coli GM2929, which is defective in the dcm dam methylation systems. The highest efficiency of electrotransformation was observed with DNA prepared from strain DH5α. The methylation level of pCD1 prepared from GM2929 was increased by in vitro modification of the DNA with the SssI methylase which methylates the dinucleotide sequence CpG (8). No difference in electrotransformation efficiency was observed under these conditions (data not shown). To further investigate the effect of DNA modification upon transformation frequencies, plasmid DNA (pMBE2) was isolated from both S. erythraea and E. coli DH5α. The plasmid DNA was used to retransform S. erythraea sonicated mycelia. There was no significant difference in transformation efficiencies between DNAs prepared from S. erythraea and E. coli DH5α. Consistent with the earlier observations of Yamamoto et al. (12), our data do not indicate that a significant restriction barrier exists for E. coli DH5α modified DNA in S. erythraea CA340.

In order to demonstrate that pMBE2 was maintained as an extrachromosomal replicon, total DNA was isolated from a hygromycin-resistant transformant, digested with any of three restriction enzymes that cleave once within pMBE2, and analyzed by agarose gel electrophoresis. No integrated plasmid was detected by hybridization of the Southern blot with a digoxigenin-labeled pMBE2 probe (Fig. 4).

FIG. 4.

Detection of transformed DNA. (A) Agarose gel of DNA samples. Lanes: 1, molecular size standards; 2 to 5, total DNA from an S. erythraea CA340 hygromycin-resistant transformant not digested and digested with BglII, HindIII, and NdeI, respectively; 6 to 9, pMBE2 from E. coli not digested and digested with BglII, HindIII, and NdeI, respectively. (B) Autoradiograph with pMBE2 as the hybridization probe.

The general utility of this method was assessed by electrotransforming other actinomycetes by using the optimized conditions for S. erythraea. pIJ702 (5) was introduced into Streptomyces lividans, and pWHM4 (11) prepared from E. coli GM2929 was introduced into Streptomyces rimosus and S. hirsuta. Although the plasmids were readily introduced into sonicated vegetative cultures, the efficiencies (2.0 × 10² to 5.0 × 10² transformants per μg of DNA) were more than 10-fold lower than those of S. erythraea. No effort was made to optimize the electrotransformation protocol for each organism, nor was the sonication dependence of organisms other than S. erythraea exhaustively examined. Nonetheless, the sonication-dependent electroporation system described here for S. erythraea appears to have general utility for the molecular genetic manipulation of this industrially important class of filamentous microorganisms.