Nov 14, 1986 - MODESTA GARCIA-DOMINGUEZ,' JUAN F. MARTIN,'* BERND MAHRO,2t ARNOLD L. DEMAIN,2. AND PALOMA LIRAS'. Departamento de ...
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1987, p. 1376-1381 0099-2240/87/061376-06$02.00/0 Copyright © 1987, American Society for Microbiology
Vol. 53, No. 6
Efficient Plasmid Transformation of the 1-Lactam Producer Streptomyces clavuligerus MODESTA GARCIA-DOMINGUEZ,' JUAN F.
MARTIN,'*
BERND MAHRO,2t ARNOLD L. DEMAIN,2 LIRAS' Departamento de Microbiologia, Facultad de Biologia, Universidad de Le6n, Le6n, Spain,' and Department of Applied Biological Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts 021392 AND PALOMA
Received 14 November 1986/Accepted 10 March 1987
The conditions for optimal formation and regeneration of protoplasts of Streptomyces clavuligerus were established. The optimal temperature for regeneration of protoplasts and for transformation was 26°C in three different regeneration media. The best efficiency of transformation was obtained with 40% polyethylene glycol 1000. The efficiencies of regeneration and transformation increased greatly when protoplasts were obtained from cultures in the early stationary phase of growth. The number of transformants per assay increased linearly with rising concentrations of protoplasts. However, the number of transformants per protoplast decreased at concentrations of protoplasts above 1.5 x 109. The total number of transformants rose linearly at increasing plasmid DNA concentrations, but the number of the transformants per microgram of DNA became constant at concentrations above 1 ,ug of DNA. Transformation frequencies as high as 5 x 105 transformants per ,ug of DNA were obtained when plasmid pU702 was isolated from S. clavuligerus but not when isolated from Streptomyces lividans. (13), and a mutant blocked in cephamycin biosynthesis (J. Romero, P. Liras, and J. F. Martin, unpublished data) were used. Cultures were kept lyophilized or as a spore suspension in 20% glycerol at -20°C. A total of 1 ml of spore suspension was inoculated into 25 ml of YEMEG medium, which contained the following per liter: 3 g of yeast extract, 5 g of peptone, 3 g of malt extract, and 10 g of glycerol (YEME medium modified by including 1% glycerol instead of glucose). The culture was incubated in triple-baffled 250-ml flasks at 30°C for the times indicated below. Protoplast formation and regeneration. Protoplasts were obtained from cultures grown at 30°C in YEMEG medium (25 ml) supplemented with sucrose (150 g/liter) and glycine (8 g/liter) under the incubation conditions described above. The mycelium was collected and washed twice with 10.3% sucrose solution, and the pellet was suspended in 2 ml of P (for protoplasting) buffer (14) containing lysozyme (1 mg/ml) and incubated at 30°C for 15 min. After dilution of the lysis solution with 2.5 ml of P buffer, the protoplast suspension was homogenized and filtered through a sterile cotton plug. The filtered protoplasts were centrifuged at 1,000 x g for 7 min and washed 3 times with P buffer and counted in a hemacytometer under a microscope. Three different protoplast regeneration media were used during this study: R5G, RSGG, and RSM. R5G was a modification of R5 (14) that contained glycerol (10 g/liter) instead of glucose. R5GG contained both glucose (10 g/liter) and glycerol (10 g/liter) and all the other components contained in R5. RSM contained the following per liter: 100 g of sucrose, 10 g of dextrin, 0.05 g of MgCl2 .6H20, 1 g of Casamino Acids (Difco Laboratories, Detroit, Mich.), 0.05 g of MgSO4 7H20, 2.5 g of L-arginine hydrochloride, 0.05 g of KH2PO4, 3.7 g of CaCl2 2H20, 5.7 g of NTris(hydroxymethyl)methyl-2-aminoethane-sulfonic acid (TES) buffer (pH 7.2), and 1 ml of trace elements. The trace elements solution contained the following per liter: 1 g of FeSO4 7H20, 1 g of MnCl2 41120, and 1 g of ZnSO4 7H20, as described by Bailey et al. (3).
Several species of Streptomyces are able to produce cephamycins and related P-lactam antibiotics (9, 19). Streptomyces clavuligerus is the most widely studied among the cephamycin producers. S. clavuligerus produces several ,-lactam antibiotics, including clavulanic acid, a potent 1-lactamase inhibitor, and cephamycin C (13, 24). Many studies have been carried out on the biosynthesis of cephamycins (2, 8, 26) and clavulanic acid (26, 27) and the control of antibiotic biosynthesis by this strain, which is becoming a model microorganism for the study of cephamycin biosynthesis. The development of recombinant DNA technology in S. clavuligerus is of great interest but has been hampered by the difficulty in regenerating cells from protoplasts and by the unavailability of suitable cloning vectors and transformation systems. A preliminary report of the cloning of DNA fragments involved in clavulanic acid biosynthesis has been published, but no detailed description of the transformation procedure was reported (3). In initial transformation experiments with S. clavuligerus by the standard procedure described for Streptomyces lividans (7, 11), we obtained only some sporadic transformants (less than 1 transformant per jg of DNA), indicating that the methods described for S. lividans are not suitable for direct application to S. clavuligerus. Several factors that affect formation of protoplasts (4, 5) and transformation of different actinomycetes (1, 21, 22) have been described. We report here the results of a study of several parameters affecting the regeneration of protoplasts and the effect of those factors and the origin of the DNA on transformation of S. clavuligerus. We obtained good transformations that allowed shotgun cloning in S. clavuligerus. MATERIALS AND METHODS Strains and culture conditions. Two strains of S. clavuligerus, a single clone of S. clavuligerus NRRL 3585 * Corresponding author. t Present address: Institut fur Genetik und Mikrobiologie, Universitat Munich, 8000 Munich LA, Federal Republic of Germany.
1376
TRANSFORMATION OF S. CLAVULIGERUS
VOL. 53, 1987
Isolation of plasmid DNA. Plasmid pIJ702 (16) carries the thiostrepton resistance determinant (tsr) from Streptomyces azureus in a pIJ350 replicon and the mel gene that encodes the tyrosinase activity involved in melanin biosynthesis (6). Katz and co-workers (16) have suggested that this plasmid has a broad host range, including S. clavuligerus. We isolated the plasmid from S. lividans or from S. clavuligerustransformed strains. Plasmid DNA from either S. lividans or S. clavuligerus was obtained by the alkaline lysis procedure (17). Plasmid was purified by cesium chloride-ethidium bromide ultracentrifugation (18). The ethidium bromide was extracted 5 times with an equal volume of isopropanol saturated with NaCl, and the DNA was precipitated with ethanol and suspended in TE buffer (10 mM Tris hydrochloride, 1 mM EDTA [pH 8]) at a concentration of 0.25 ,ug/,ul. Agarose gel electrophoresis and restriction endonuclease analysis were carried out by standard methods (18). Transformation of protoplasts. The basic transformation procedure was as follows. Protoplasts were prepared as described above (about 2.5 x i07 protoplasts per experiment), centrifuged, and suspended in 100 iil of P buffer. In some experiments salmon sperm DNA (0.8 pug per assay) and protamine sulfate (1.5 ,ug per assay; Sigma Chemical Co., St. Louis, Mo.) were added at this stage. The corresponding amount of pIJ702 DNA (0.01 to 4 pRg) from S. clavuligerus, unless indicated otherwise, was then added in 20 pl of TE buffer and mixed. Immediately after this, 0.5 ml of polyethylene glycol (PEG) in P medium was added and mixed by gentle shaking. After about 1 min at room temperature, a dilution was made by adding 2.5 ml of P buffer, and the protoplasts were collected by centrifugation at 2,000 x g and suspended in 1 ml of P medium. Dilutions were made quickly, and the transformed protoplasts were plated (0.1 ml per plate) in R5G, R5GG, and RSM media. Agar plates were partially dehydrated by drying them in a laminar flow cabinet for at least 2 h (5). The plates were then incubated at temperatures ranging from 26 to 30°C. After about 36 h (depending on the regeneration medium), when tiny colonies started to appear, the plates were overlaid with 2.5 ml of DNA soft agar (14) containing 8 ,ug of thiostrepton per ml to give a final concentration of 1 ,ug/ml after equilibration with the bottom layer of agar. Thiostrepton-resistant clones were counted after 3 to 4 days (RSG or RSGG media) or 5 to 6 days (RSM medium). Transformant clones were transferred to thiostrepton-containing plates and identified as S. clavuligerus by the inability to utilize glucose (2, 13, 26) and the production of cephamycin C and clavulanic acid. The plasmid was isolated from transformant clones and shown to be identical to the transforming vector by restriction endonuclease analysis. Chemicals. PEG 1000 was obtained from Sigma, KochLight Ltd. (Haverhill, Suffolk, United Kingdom), or MerckSchuchardt (Munich, Federal Republic of Germany). PEG 6000 was obtained from Merck-Schuchardt. Thiostrepton was kindly provided by S. J. Lucania of E. R. Squibb and Sons, New Brunswick, N.J. Restriction endonucleases were purchased from New England BioLabs, Inc., Beverly, Mass.; Bethesda Research Laboratories, Gaithersburg, Md.; or Amersham Laboratories, Buckinghamshire, England. All other chemicals were of reagent quality. RESULTS Optimization of protoplast formation and regeneration at increasing concentrations of thiostrepton. S. clavuligerus became sensitive to lysozyme after growth at 28°C in
1377
A
~~~~z
~ ~
~
~
1., z
w
3,
-
RSM RSG RS(;G RSM RSG RSGG FIG. 1. (A) Regeneration of protoplasts in RSM, M5G, or RSGG medium at 26°C (O) or 300C (-). (B) Transformation of protoplasts in the same media at 26°C (O) or 300C (-). Protoplasts were obtained from cultures that were grown for 45 h and transformed with DNA from S. clavuligerus as indicated in the text in the absence of ATA and without heating of the protoplasts.
YEMEG medium supplemented with glycine (8 g/liter) and sucrose (150 g/liter). Concentrations of glycine of 2 and 4 g/liter did not affect growth, but concentrations of 6 and 8 g/liter reduced the growth rate of the culture by 50%. The optimal sucrose concentration for protoplast release was found to be 15%. When sucrose was increased to 20 or 34%, growth was completely inhibited. More than 95% of mycelia grown in YEMEG supplemented with glycine (8 g/liter) and sucrose (150 g/liter) were converted into protoplasts after only 15 min of lysozyme (1 mg/ml) treatment. The regeneration of control, untransformed protoplasts of S. clavuligerus NRRL 3585 was studied in the presence of increasing concentrations of thiostrepton. An interesting finding was the great sensitivity of S. clavuligerus to thiostrepton. The MIC of thiostrepton under conditions of protoplast regeneration was about 0.5 ,ug/ml; and no regeneration at all took place at 1, 5, and 25 jig/ml. Therefore, the selective final concentration of thiostrepton used in the protoplast regeneration medium was routinely 1 ,ug/ml, although transformants could be replicated and grown perfectly at concentrations of thiostrepton as high as 20 ,ug/ml. Exactly the same MIC was obtained when the nce-2 mutant, blocked in cephamycin biosynthesis, was used as the host in transformation experiments; and an identical level of resistance to thiostrepton was observed after transformation with
pIJ702. Effect of temperature and medium composition on protoplast regeneration. There was an important effect of the temperature at which protoplasts were regenerated on the efficiency of regeneration (Fig. 1A). The optimal temperature for regeneration of protoplasts of several Streptomyces isolates was 29 to 30°C (5, 14). In S. clavuligerus, however, the efficiency of regeneration increased substantially when the temperature was decreased from 30 to 26°C in the three different media that were used. The effect of decreasing the temperature to 26°C was particularly evident in RSM medium (about a sevenfold increase in regeneration) as compared with RSG or R5GG medium. Protoplast regeneration in RSM medium was very slow, however (requiring 108 to 132 h), as compared with that in RSG or RSGG medium (72 to 96 h). Modifications of the RSM medium were carried out to try to get a faster regeneration of protoplasts. Substitution
1378
APPL. ENVIRON. MICROBIOL.
GARCIA-DOMINGUEZ ET AL.
of maltose, glycerol, glutamic acid, or proline (other carbon sources that are used efficiently by S. clavuligerus) for the dextrin in RSM medium did not speed up protoplast regeneration. Therefore, other parameters were optimized in R5G medium. The best efficiency of regeneration was obtained by using TES buffer in the regeneration media. When 3-(Nmorpholino)-propanesulfonic acid (MOPS) buffer was used at the same concentration and pH, the efficiency of regeneration decreased two- or threefold. An increase in the number of protoplasts that were used per plate had a slightly negative effect on protoplast regeneration, but no clear autoinhibition of nearby regenerating protoplasts was observed, as has been described previously for Streptomyces fradiae (5) or Streptomyces coelicolor (15). A 102 to 104 dilution of the protoplast suspension after transformation, however, was convenient for selecting individual transformant clones. Effect of media and temperature during regeneration on protoplast transformation. The efficiency of transformation was higher in RSM than in R5G or R5GG media (Fig. 1B), although transformants took 1 to 2 additional days to grow in this medium, as was described above for protoplast regeneration. The temperature of regeneration of protoplasts had only a small effect on transformation (as compared with the large effect on protoplast regeneration), suggesting that factors other than regeneration (e.g., DNA uptake and replication; see below) became limiting for transformation. The best transformation was observed at 26°C. Effect of the type and concentration of PEG on transformation. The effect of different concentrations of PEG 1000 and PEG 6000 from three commercial sources on transformation of S. clavuligerus protoplasts is shown in Fig. 2. PEG 1000 (in the range of 20 to 60%) consistently gave a much higher efficiency of transformation than PEG 6000 in several experiments. There were large differences in the effect of PEG from different sources. The best results were obtained with PEG 1000 from Koch-Light. PEG 1000 from Merck gave intermediate results, whereas PEG 1000 from Sigma gave very poor results. High concentrations of PEG 1000 (40% or above) gave the highest transformation frequencies.
z 0
z 'i
0
I.-
0)
0
20
40 60 PEG (o/*) FIG. 2. Effect of PEG type and concentration on the frequency of transformation. Protoplasts were obtained from a 45-h culture, transformed, and regenerated in R5G medium at 26°C. PEG 1000 from Merck (0), PEG 6000 from Merck (0), PEG 1000 from Koch-Light (A), and PEG 1000 from Sigma (K) were used. The transformation conditions were as described in the legend to Fig. 1.
~~~~~~~~~~~c 20 40 60 80 1 200
C
0
-.--
cr Z~~~~~~~~~~~~w
6
c0w
0 LL
0. 100
0
cr~~~~~~~~~~~~~~
r
a.
0
20
40
60 TIME(h)
80
100
FIG. 3. Effect of S. clavuligerus cell growth phase (A) on protoplast regeneration in R5G medium (0) and transformation (0). The transformation conditions were as described in the legend to Fig. 1 (26°C). OD, Optical density.
Effect of cell growth phase on protoplast regeneration and transformation. The growth phase of the culture had a large influence on the efficiencies of protoplast regeneration and transformation. The efficiency of protoplast regeneration increased very clearly after the exponential phase of growth, reaching a maximum value at 60 h (Fig. 3). About 35% of the protoplasts counted microscopically regenerated at this time. The regeneration efficiency decreased drastically when protoplasts were obtained from stationary-phase cultures. The efficiency of transformation correlated rather well with the ability of the protoplasts to regenerate (Fig. 3). The ratio of transformants to viable protoplasts present in the transformation mixture before the addition of PEG (i.e., able to regenerate) was relatively constant over the cell density range that occurred during growth of the culture. Effect of the concentration of protoplasts on transformation. When increasing concentrations of protoplasts (in the range of 1 x 108 to 3 x 109) were used, the number of transformants per assay (always with 1 ,g of DNA) increased linearly (Fig. 4). When the number of transformants per regenerated protoplast was calculated, however, the value increased up to 4 x 108 protoplasts and then remained relatively constant up to protoplast concentrations of 1.5 x 109. Higher concentrations of protoplasts (3 x 109) gave a lower efficiency of transformation per protoplast. In all cases, only a minor fraction of protoplasts (less than 1 in 104) was susceptible to transformation even at a high DNA concentration. Effect of plasmid DNA and carrier DNA on transformation. Increasing concentrations of pIJ702 were used in this experiment. The total number of transformants increased linearly as the plasmid DNA concentration in the transformation assay was raised from 0.04 to 4 p,g (Fig. 5). In this experiment, 3 x 109 protoplasts were used to avoid interference by a limited number of protoplasts. When the number of transformants per microgram of input DNA was calculated, however, the transformation efficiency increased up to 1 ,ug of total DNA and then became approximately constant. The addition of heterologous carrier DNA (salmon sperm) at a concentration of 1.5 ,ug per assay, protamine sulfate (a DNA binding agent) at 0.8 jig per assay, or both did not increase the transformation efficiency unlike the situation in S. fradiae and Streptomyces ambofaciens (21). In fact, the
1379
TRANSFORMATION OF S. CLAVULIGERUS
VOL. 53, 1987
tn
4
-
4
4 LA
LA 0. Ln
4':
I.-
z
I z LA:
LA z 4 cc
0 I-.
Ifcr
in
Q
0
10
20
108 PROTOPLASTS FIG. 4. Effect of the protoplast concentration on the efficiency of , Total number of transformants per assay; transformation. number of transformants per protoplast. The transformation conditions were as described in the legend to Fig. 1 (26°C). Protoplasts were obtained from cells that were grown for 60 h.
addition of protamine sulfate decreased slightly the efficiency of transformation. Influence of the origin of plasmid DNA. A large difference in transformation efficiencies was observed when plasmid pIJ702 was obtained from S. clavuligerus or S. lividans. Using several batches of pIJ702 from S. lividans, we obtained less than 1 transformant per jig of DNA, whereas transformation efficiencies as high as 5 x 105 transformants per ,g of DNA were obtained with the homologous plasmid from S. clavuligerus. These results suggest the presence of a restriction system in S. clavuligerus. Homologous pIJ702 (from S. clavuligerus) transformed both the wild type S. clavuligerus NRRL 3585 and the nce-2 mutant with similar efficiencies. Effect of nuclease inhibitors on transformation. Attempts were made to remove nonspecific DNases by heating the protoplasts at 30°C for 15 to 30 min and by the addition of aurintricarboxylic acid (ATA), a known inhibitor of nucleases (23), at concentrations of 0.01 to 0.1 mM. Both treatments were detrimental for cell viability. Heating of the protoplasts for 30 min reduced the viability of the protoplasts by 70%; concentrations of ATA higher than 0.1 mM were highly inhibitory for protoplast regeneration even after the protoplasts were washed five times to remove the dye. A slight increase in the efficiency of protoplast transformation was obtained by using these two approaches. Heating of the protoplasts for 15 min at 30°C (in the absence of ATA) increased the transformation efficiency 1.5- to 2-fold, and the addition of ATA (after heating for 15 min at 30°C) increased the efficiency slightly (1.9 x 105 at 0.01 mM and 2.2 x 105 at 0.05 mM). This increase was of little relevance, however, because similar efficiencies of transformation were obtained in some experiments without the addition of ATA. DISCUSSION
The extremely low initial transformation frequency of S. clavuligerus with plasmid DNA isolated from S. lividans
excluded the possibility of direct cloning in this strain by complementation of nonproducing mutants. A major improvement in the efficiency of transformation of S. clavuligerus was obtained by optimizing several parameters that affect protoplast regeneration, transformation, or both. Both strains of S. clavuligerus were more sensitive to thiostrepton (MIC, about 0.5 ,ug/ml) under protoplast regeneration conditions than S. lividans (14) and other Streptomyces species, e.g., S. ambofaciens (21). Moreover, the thiostrepton resistance marker carried by pIJ702 (6, 16) was well expressed in S. clavuligerus, thus allowing good selection of transformants by resistance to this antibiotic. The optimal temperature for protoplast regeneration and transformation of S. clavuligerus was 26°C, as indicated also by Bailey et al. (3). This was about 3 to 4°C lower than the optimal temperature (29 to 30°C) for the transformation of other Streptomyces species (5, 7, 29). Regeneration and transformation of S. clavuligerus protoplast was much better in the RSM medium containing dextrin, sucrose, and arginine as carbon sources than in R5G or R5GG medium. However, RSM medium was rather unsuitable due to the long time (108 to 132 h) required for protoplast regeneration. S. clavuligerus is a peculiar microorganism in that it is unable to utilize glucose, but it grows well in glycerol (13, 26). Protoplast regeneration was more rapid (72 to 96 h) in R5G or R5GG medium containing glycerol, although the efficiencies of regeneration in these media were lower. Because the large increase in protoplast regeneration that was observed when the protoplasts were regenerated at 260C as compared with 30°C was not followed by a similar increase in transformation (compare Fig. 1A and B), it seems that the efficiency of transformation is limited by factors other than regeneration; e.g., it may be limited by DNA uptake and replication in the viable protoplasts. This hypothesis is supported by the observation that only a small fraction of total protoplasts was able to replicate and express the markers of the introduced plasmid. A similar conclusion
10.0
5.0 4 z
4 %n
0 In
1-0
z
z 4
1 0.5
%n
VI) z
2 F 0.10
0
0.05
II
I
0.05
0.1
0.5
1.0
