- Email: [email protected]
Transformation of Bacillus subtilis in Chocolate Milk: Evidence for Low Frequency of Establishment of Cells Transformed Under Non-selective Conditions
System. Appl. Microbiol. 25, 478–482 (2002) © Urban & Fischer Verlag http://www.urbanfischer.de/journals/sam
Transformation of Bacillus subtilis in Chocolate Milk: Evidence for Low Frequency of Establishment of Cells Transformed Under Non-selective Conditions Anja Wittke, Sonja Lick, and Knut J. Heller Institute for Microbiology, Federal Dairy Research Center, Kiel, Germany Received: August 15, 2002
Summary Transformation of naturally competent Bacillus subtilis with plasmid was carried out in chocolate milk without antibiotics. Transformed cells were enumerated during the entire growth phase in chocolate milk. When DNA was added to aliquots of a batch culture after different times of incubation, transformation events were detected at all different growth stages. When DNA was added to a batch culture together with the inoculum, transformed cells were detected at the onset of exponential growth. However, apparently no or only limited growth of these transformed cells was observed. To clarify, whether the limitation of growth was due to suppression by non-transformed cells, different proportions of B. subtilis cells either carrying or not carrying the plasmid were mixed and inoculated into chocolate milk without antibiotic. Our results indicate that suppression appears to be of minor importance. Instead, plasmidbearing cells appear to suffer from a prolonged lag-phase. However, the failure to exhibit significant growth of cells which had taken up the plasmid in chocolate milk appears to be due to failure of these cells to establish themselves as permanently transformed under non-selective conditions. Key words: Genetically modified DNA – transformation – competence – Bacillus subtilis – milk product
Introduction One of the concerns when applying genetic engineering in food biotechnology is that the new genetic traits introduced may spread to other organisms by horizontal gene transfer, either in the food itself, or in the intestine of the consumer, or after release from the intestine of the consumer in the sewage plant or in the soil. Different transfer scenarios may be imagined. One scenario would involve genetically modified microorganisms used e.g. as starter cultures [10]. They may transfer their recombinant DNA either directly by conjugation or transduction, or they may transfer it indirectly by releasing their DNA into the food. Another scenario would involve DNA being present in food additives as a result of the lysis of the genetically modified organisms producing the additive. For the latter two mechanisms involving free, naked DNA, naturally competent bacteria [8] would be needed to take up the DNA from the food. Such organisms may be either general constituents of a fermentation flora or contaminating organisms. In this communication we analyzed in a model system the uptake of free DNA in a food stuff. Several studies
concerning this aspect have been carried out recently [3, 11, 2]. As the bacterial model Bacillus subtilis had been chosen. B. subtilis, including the “natto” strain used for fermentation of soybeans [1], is known for its potential for natural transformation. Plasmid pMG36enpr [9] was used as the DNA model. The plasmid replicates in lactococci as well as in B. subtilis, has a size of 6.0 kb, and confers erythromycin resistance (EmR). It may thus be used as a “worst case” model allowing the detection of very low transfer frequencies. As a food model chocolate milk was chosen, as it has recently been shown that this food stuff allows the development of competence in B. subtilis [11].
Material and Methods Bacterial strains and growth conditions B. subtilis 168 (trpC) [5] was propagated either in Nutrient broth No. 2 (NB-2) medium (Oxoid, Hampshire, GB), in chocolate milk (1.8% fat contents, UH-treated, Hansano®) or in MC(–)-medium [11] at 37 °C with vigorous shaking in fourfold 0723-2020/02/25/04-478 $ 15.00/0
Transformation of Bacillus subtilis in Chocolate Milk baffled Erlenmeyer flasks. Transformed cells were plated on solidified NB-2 medium containing erythromycin (5 µg/ml) and lincomycin (50 µg/ml). DNA extraction and preparation Plasmid DNA was prepared using a plasmid preparation kit (Macherey & Nagel, Düren, FRG). Cells resuspended in buffer A1 were first digested with lysozyme and incubated for 30–60 min at 37 °C. All further steps were performed as described by the manufacturer including the step using buffer A3. To the supernatant 20–50 µl of proteinase K (20 mg/ml) was added and the solution was incubated for 15–30 min at 37 °C. It was extracted with an equal volume of phenol/chloroform (1:1) and chloroform/isoamyalcohol (24:1), before it was loaded onto the column. To quantify plasmid DNA obtained, plasmids were digested with a restriction enzyme cutting only once, separated by agarose gel electrophoresis, and compared with a suitable quantitation mass ladder after ethidium bromide staining. Measurement of DNA amounts was performed using the image analysis program Sigma Scan Pro 5 (Sigma-Aldrich, Taufkirchen, FRG). Transformation experiments in chocolate milk B. subtilis was propagated overnight in chocolate milk. Fresh chocolate milk was inoculated with 0.05% of the overnight culture. DNA was added in two different ways: i) the total amount of plasmid DNA (0.1 µg/ml) was added together with the inoculum to an 8 ml culture; ii) to 200 µl aliquots taken in hourly intervals from the growing culture (8ml), plasmid DNA (0.1 µg/ml) was added. Aliquots were incubated 60 min at 37 °C with vigorous shaking. As a control, the transformation was performed in parallel in MC(-) medium. Growth of varying proportions of pMG36enpr-harbouring and plasmid-free B. subtilis Overnight cultures of B. subtilis 168 or B. subtilis 168(pMG36enpr) grown in NB-2 were adjusted to an equal optical densities and used to inoculate chocolate milk. The proportions of plasmid-free to plasmid-containing B. subtilis inocula were in the ratios 1:1, 1:0.1, 1:0.01, 1:0.001. Different dilutions of the cultures were plated on solidified NB-2 medium (for determination of total cell counts) as well as on NB-2 medium containing erythromycin (5 µg/ml) and lincomycin (50 µg/ml) (for determination of plasmid-bearing cell counts).
479
development of competence, aliquots were removed from the growing culture each hour, starting immediately after addition of the inoculum. pMG36enpr was added followed by incubation for 60 min at 37 °C with vigorous shaking. Thereafter, the mixture was plated onto NB-2Em agar plates. Plates were incubated overnight at 37 °C. A low number of competent cells in the growing culture was first detected after 10 hours. The number increased during the next 3 to 4 hours in the stationary phase and then started to decrease until termination of the experiment after 16 hours (Fig. 1). Highest transformation rates observed were ca. 6 × 103/µg DNA. When B. subtilis was grown in chocolate milk a very different kinetic of development of competence was observed. First we noted that growth rate as well as cell yield were higher than in MC(-)-medium, indicating that chocolate milk was apparently a complex and rather good growth medium for B. subtilis. It was, however, a suitable medium for development of competence. This had already been observed by Bräutigam et al. [3] and Zenz et al. [11]. Competent cells, the presence of which was again tested for by removal of aliquots from the growing culture and addition of pMG36enpr, were identified throughout all stages of growth (Fig. 2a). Fluctuations in the number of competent cells were observed with three peaks immediately after inoculation, at the end of exponential growth, and after several hours into stationary phase, respectively. For the peak immediately after inoculation carry-over of competent cells from the stationary over-night culture seems to be most likely. The highest transformation rate detected was 6 × 103/µg DNA after ca. 6 h into stationary phase. It should be mentioned that – although some of the absolute numbers of colony forming units on the agar plates were rather low -– comparable results were obtained in all four repetitions of the experiment. In order to apply a more realistic model concerning the uptake of DNA from food stuff by B. subtilis, we added pMG36enpr to the culture once together with the B. sub-
Quantitation of pMG36enpr DNA by polymerase chain reaction (PCR) During the transformation experiment samples were taken every hour, frozen, and later analyzed for the amount of pMG36enpr DNA. Different fragments of the 6.0 kb pMG36enpr plasmid were amplified by PCR from different dilutions of the plasmid as described by Lick et al. [7], to determine the detection limit for each fragment. For the amplification of a 4.2 kb fragment [7], the detection limit was 0.1 µg/ml in chocolate milk.
Results and Discussion Bacillus subtilis growing in batch culture has been described to develop competence during transition from exponential to stationary growth phase only. As one prerequisite for development of competence, growth in minimal salts instead of complex medium has been identified [4, 6]. We could confirm these results by growing B. subtilis in MC(–)-medium [11]. Exponential growth was observed between 2 and 10 h after inoculation. To test for
Fig. 1. Development of competence of B. subtilis in MC(–)medium. Solid symbols: viable counts (cfu/ml); open symbols: transformed cells counts (cfu/µg DNA).
480
A. Wittke et al.
tilis inoculum. In this case aliquots were removed from the culture and immediately plated onto NB-2-Em agar plates. One hour after inoculation the number of EmR cells was almost exactly the same as in the experiment where aliquots had been withdrawn (Fig. 2b). However, in contrast to our expectations, the number of EmR cells did not increase in parallel to the number of total cells of the culture. Instead, the number of EmR cells decreased within the next two hours, and then slightly increased until the end of the experiment at 24 h after inoculation. The final number of EmR cells at 24 h was ca. 8 × 103/µg DNA. If growth parallel to the total number of cells of the culture would have occurred, the number would have reached ca. 106/µg after 24 h. One interpretation could be that plasmid-bearing cells being present in low amounts
in a culture of otherwise plasmid-free cells suffer from suppression by the plasmid-free cells under non-selective conditions. To test for the latter possibility, different proportions of plasmid-bearing and plasmid-free cells were inoculated from overnight cultures together into chocolate milk at 37 °C. The experiment was carried out twice with almost identical outcome. Growth was monitored by plating different dilutions of the cultures on agar plates with Em (for plasmid-bearing cells) or without Em (for total cells) (Fig. 3a–e). From the samples taken immediately after inoculation the following proportions (total cells:plasmidbearing cells) were calculated: 1:0.38 (Fig. 3a), 1:0.027 (Fig. 3b), 1:0.002 (Fig. 3c), 1:0.00014 (Fig. 3d), 1:1 (Fig. 3e; control 1, plasmid-bearing cells only), 1:0 (Fig. 3f;
Fig. 2. Development of competence of B. subtilis in chocolate milk. a: DNA added to aliquots; b: DNA added once together with the inoculum; solid symbols: viable counts(cfu/ml); open symbols: transformed cells counts (cfu/µg DNA).
Fig. 3. Growth of pMG36enprharbouring and plasmid-free B. subtilis cells, inoculated in varying proportions (plasmid-free cells:plasmid-bearing cells) from overnight cultures. a: 1:0.38; b: 1:0.027; c: 1:0.002; d: 1:0.00014; e: plasmid-bearing cells alone; f: plasmid-free cells alone. Solid symbols: total cells; open symbols: plasmid-bearing cells.
Transformation of Bacillus subtilis in Chocolate Milk
control 2, plasmid-free cells only). It should be noted, that in all cases growth in chocolate milk was in the absence of Em. Several observations were made. i) Growth of plasmid-bearing cells was always slower than corresponding growth of the total cultures. The increase of cell numbers of plasmid-bearing cells within the first four hours was by factors between ca. 25 and 39. This was also true for the culture consisting of plasmid-bearing cells only (Fig. 3e). In contrast, total numbers of cells increased within the first four hours by factors of ca. 150 to 400, with 400 being the increase for the culture consisting of plasmid-free cells only. ii) The increase in plasmidbearing cell numbers in mixed cultures (Fig. 3a–d) was highest for the culture with the highest proportion of plasmid-bearing cells (Fig. 3a). The increase during the first 16 h in the latter case was by a factor of 4.6 × 103, while it was between 6.4 and 8.5 × 102 during the first 12 (Fig. 3b) or 16 h (Fig. 3c and d) for the other three cultures. The slightly delayed growth of plasmid-bearing cells in relation to total (mostly plasmid-free) cells as described under i) may partly be responsible for the reduced multiplication of the plasmid-bearing cells during the first 4 h, since the growth rate of the cells within a culture is determined by the growth rate of the majority of cells. Thus, the plasmid-bearing cells in a mixed culture were not able to achieve their highest growth rate between 4 and 8 h, since the culture as a whole was already beyond is maximal growth rate. iii) The highest absolute cell numbers reached at one point during the 48 h measuring period were comparable for all six cultures, being between 3 and 8 × 109 per ml. Theses numbers were reached, however, at different incubation times: 16 h for the cultures shown in Fig. 3a and 3e, 24 h for those shown in Fig. 3b, 3c, an 3d, and 48 h for growth of plasmid-free cells as shown in Fig. 3f. iv) As a consequence of the observations in iii), all cultures except the one consisting of plasmid-free cells only showed a decline of the absolute cell numbers during stationary phase. The decline was highest in the culture consisting of plasmid-bearing cells only, resulting in a reduction by more than two logs. In all the mixed cultures the number of plasmid-bearing cells declined by less than two logs. It should be noted that in none of the mixed cultures the reduction of the number of plasmid-bearing cells would be sufficient to account for the reduction of the total cell numbers. v) Growth of plasmid-bearing cells in chocolate milk under non-selective conditions did not result in a detectable loss of the plasmid (Fig. 3e). The differences in cell numbers obtained from colony counts on plates with or without Em fell within the regular variations (± 30%) observed for this technique. On the basis of the results obtained with the mixed cultures, the experiment shown in Fig. 2b may be interpreted as follows. i) Uptake of pMG36enpr DNA into the cells appears to take place during the first hour only. Thereafter, the DNA in the medium may become inaccessible for transformation by association with material present in chocolate milk, such material being e.g. casein. Degradation can be excluded for the first 6 h, since the DNA was detectable by PCR at a level of 0.1 µg/ml dur-
481
ing this period (data not shown). ii) Lack of growth of the EmR cells detectable after 1 h can neither be explained by the suppressing activities of the plasmid-free cells, which would only results in a decreased growth rate but not in total inhibition of growth, nor by the lag-phase observed for growth of plasmid-bearing in contrast to plasmid-free cells. Only if this lag-phase was at least three to four hours after uptake of pMG36enpr, the growth curve of the EmR cells could be explained. iii) We propose that the kinetics of appearance of EmR cells is the consequence of two different kinetics: the first being represented by the loss of pMG36enpr under non-selective conditions from cells shortly after uptake of the plasmid, and the second being represented by the growth of a small fraction of EmR cells which have established themselves as permanently pMG36enpr-transformed cells under non-selective conditions. This would mean that of the ca. 5 × 102 EmR cells present after 1 h only about 5 would become permanently transformed and would be responsible for the growth curve observed after the fourth hour after inoculation.
Conclusions One of the hazards described for genetic engineering of bacteria is the transfer of the recombinant DNA into other bacteria of the environment. Especially if recombinant plasmids are used the question will not be whether transfer does occur but to what extend it does occur. The results of this communication allow some quantitative calculations to be made for our model system. When 0.1 µg recDNA and 106 B. subtilis cells are present in one ml of chocolate milk, about one of each 2000 cells will take up DNA. Of these, one of each 100 cells will become permanently transformed. Upon growth, these cells will be lagging by about one log behind the plasmid-free cells. In this model, two numbers have to be questioned. The amount of 0.1 µg recDNA, if considered to be introduced into the chocolate milk e.g. via an additive isolated from a genetically engineered microorganism, is very high. Treatment with nuclease would easily lower the amount to less than 0.1 ng/ml. This would lower the number of transformants concomitantly. The number of 106 B. subtilis cells per ml is also very high. Milk, contaminated with such an amount of saprophytic cells, would already be considered “spoilt”, it would no longer be acceptable for consumption. A contamination e.g. at the process of aseptic filling would result in a likely contamination with only a few cells per milk container with a volume of 1 l. Extrapolating our transformation numbers to a chocolate milk containing 0.1 ng recDNA and 10 B. subtilis cells per ml, 108 ml – corresponding to 100 tons of chocolate milk – would be needed to give rise to 5 permanently transformed cells. Needless to say that even this low number of transformation events is still overestimated i) because of the linear extrapolation, ii) because of the still rather high number assumed for contaminating B. subtilis cells, and iii) because of the assumption that pure plasmid DNA was introduced via an additive. Under realistic con-
482
A. Wittke et al.
ditions, the plasmid would only constitute a minor fraction of the assumed amount of 0.1 ng DNA present per ml.. The majority of the non-plasmid, non-recDNA could act as a competitor for plasmid-uptake, reducing the chances for plasmid transformation even further. Acknowledgements We like to express our thanks to Katrin Schneede and Meike Groszek for excellent technical assistance. This work was in part supported by the German Federal Ministry for Education, Science, Research, and Technology (Project No. 0311049 and 0311049A).
6. 7.
8. 9.
References 1. Ashikaga, S., Nanamiya, H., Ohashi, Y., Kawamura, F.: Natural genetic competence in Bacillus subtilis natto OK2. J. Bacteriol. 182, 2411–2415 (2000). 2. Bauer, F., Hertel, C., and Hammes, W. P.: Transformation of Escherichia coli in foodstuffs. Syst. Appl. Microbiol. 22, 161–168 (1999). 3. Bräutigam, M., Hertel, C., and Hammes, W. P.: Evidence for natural transformation of Bacillus subtilis in foodstuffs. FEMS Microbiol. Lett. 155, 93–98 (1997). 4. Dubnau, D. and Roggiani, M.: Growth medium-independent genetic competence mutants of Bacillus subtilis. J. Bacteriol. 172, 4048–4055 (1990). 5. Errington, J. and Mandelstam J.: Use of a lacZ gene fusion to determine the dependence pattern of sporulation operon
10.
11.
spoIIA in spo mutants of Bacillus subtilis. J. Gen. Microbiol. 132, 2967–2976 (1986). Hauser, P. M. and Karamata, D.: A rapid and simple method for Bacillus subtilis transformation on solid media. Microbiology 140, 1613–1617 (1994). Lick, S., Schneede, K., Wittke, A., Heller, K. J.: Detection without sample processing of yoghurt starters and DNA in milk products by PCR. Kieler Milchwirtschaftl. Forsch.ber. 51: 15–25 (1999). Lorenz, M. G. and Wackernagel, W.: Bacterial gene transfer by natural genetic transformation in the environment. Microbiol. Rev. 58, 563–602 (1994). van de Guchte, M., Kodde, J., van der Vossen, J. M., Kok, J., and Venema, G.: Heterologous gene expression in Lactococcus lactis subsp. lactis: synthesis, secretion, and processing of the Bacillus subtilis neutral protease. Appl. Environ. Microbiol. 56, 2606–2611 (1990). von Wright, A. and Sibakov, M.: Genetic Modification of Lactic Acid Bacteria. In „Lactic Acid Bacteria, Microbiology and Functional Aspects“ (S. Salminen and A. von Wright, Eds.), pp. 161–210. Marcel Dekker Inc. New York (1998). Zenz, K. I., Neve, H., Geis, A., and Heller, K. J.: Bacillus subtilis develops competence for uptake of plasmid DNA when growing in milk products. Syst. Appl. Microbiol. 21, 28–32. (1998).
Corresponding author: Knut J. Heller, Institute for Microbiology, Federal Dairy Research Center, Hermann-Weigmann-Str.1, 24103 Kiel, Germany Tel.: ++49-(0)431-609-2340; Fax: ++49-(0)431-609-2306; e-mail: [email protected]
Transformation of Bacillus subtilis in Chocolate Milk: Evidence for Low Frequency of Establishment of Cells Transformed Under Non-selective Conditions Anja Wittke, Sonja Lick, and Knut J. Heller Institute for Microbiology, Federal Dairy Research Center, Kiel, Germany Received: August 15, 2002
Summary Transformation of naturally competent Bacillus subtilis with plasmid was carried out in chocolate milk without antibiotics. Transformed cells were enumerated during the entire growth phase in chocolate milk. When DNA was added to aliquots of a batch culture after different times of incubation, transformation events were detected at all different growth stages. When DNA was added to a batch culture together with the inoculum, transformed cells were detected at the onset of exponential growth. However, apparently no or only limited growth of these transformed cells was observed. To clarify, whether the limitation of growth was due to suppression by non-transformed cells, different proportions of B. subtilis cells either carrying or not carrying the plasmid were mixed and inoculated into chocolate milk without antibiotic. Our results indicate that suppression appears to be of minor importance. Instead, plasmidbearing cells appear to suffer from a prolonged lag-phase. However, the failure to exhibit significant growth of cells which had taken up the plasmid in chocolate milk appears to be due to failure of these cells to establish themselves as permanently transformed under non-selective conditions. Key words: Genetically modified DNA – transformation – competence – Bacillus subtilis – milk product
Introduction One of the concerns when applying genetic engineering in food biotechnology is that the new genetic traits introduced may spread to other organisms by horizontal gene transfer, either in the food itself, or in the intestine of the consumer, or after release from the intestine of the consumer in the sewage plant or in the soil. Different transfer scenarios may be imagined. One scenario would involve genetically modified microorganisms used e.g. as starter cultures [10]. They may transfer their recombinant DNA either directly by conjugation or transduction, or they may transfer it indirectly by releasing their DNA into the food. Another scenario would involve DNA being present in food additives as a result of the lysis of the genetically modified organisms producing the additive. For the latter two mechanisms involving free, naked DNA, naturally competent bacteria [8] would be needed to take up the DNA from the food. Such organisms may be either general constituents of a fermentation flora or contaminating organisms. In this communication we analyzed in a model system the uptake of free DNA in a food stuff. Several studies
concerning this aspect have been carried out recently [3, 11, 2]. As the bacterial model Bacillus subtilis had been chosen. B. subtilis, including the “natto” strain used for fermentation of soybeans [1], is known for its potential for natural transformation. Plasmid pMG36enpr [9] was used as the DNA model. The plasmid replicates in lactococci as well as in B. subtilis, has a size of 6.0 kb, and confers erythromycin resistance (EmR). It may thus be used as a “worst case” model allowing the detection of very low transfer frequencies. As a food model chocolate milk was chosen, as it has recently been shown that this food stuff allows the development of competence in B. subtilis [11].
Material and Methods Bacterial strains and growth conditions B. subtilis 168 (trpC) [5] was propagated either in Nutrient broth No. 2 (NB-2) medium (Oxoid, Hampshire, GB), in chocolate milk (1.8% fat contents, UH-treated, Hansano®) or in MC(–)-medium [11] at 37 °C with vigorous shaking in fourfold 0723-2020/02/25/04-478 $ 15.00/0
Transformation of Bacillus subtilis in Chocolate Milk baffled Erlenmeyer flasks. Transformed cells were plated on solidified NB-2 medium containing erythromycin (5 µg/ml) and lincomycin (50 µg/ml). DNA extraction and preparation Plasmid DNA was prepared using a plasmid preparation kit (Macherey & Nagel, Düren, FRG). Cells resuspended in buffer A1 were first digested with lysozyme and incubated for 30–60 min at 37 °C. All further steps were performed as described by the manufacturer including the step using buffer A3. To the supernatant 20–50 µl of proteinase K (20 mg/ml) was added and the solution was incubated for 15–30 min at 37 °C. It was extracted with an equal volume of phenol/chloroform (1:1) and chloroform/isoamyalcohol (24:1), before it was loaded onto the column. To quantify plasmid DNA obtained, plasmids were digested with a restriction enzyme cutting only once, separated by agarose gel electrophoresis, and compared with a suitable quantitation mass ladder after ethidium bromide staining. Measurement of DNA amounts was performed using the image analysis program Sigma Scan Pro 5 (Sigma-Aldrich, Taufkirchen, FRG). Transformation experiments in chocolate milk B. subtilis was propagated overnight in chocolate milk. Fresh chocolate milk was inoculated with 0.05% of the overnight culture. DNA was added in two different ways: i) the total amount of plasmid DNA (0.1 µg/ml) was added together with the inoculum to an 8 ml culture; ii) to 200 µl aliquots taken in hourly intervals from the growing culture (8ml), plasmid DNA (0.1 µg/ml) was added. Aliquots were incubated 60 min at 37 °C with vigorous shaking. As a control, the transformation was performed in parallel in MC(-) medium. Growth of varying proportions of pMG36enpr-harbouring and plasmid-free B. subtilis Overnight cultures of B. subtilis 168 or B. subtilis 168(pMG36enpr) grown in NB-2 were adjusted to an equal optical densities and used to inoculate chocolate milk. The proportions of plasmid-free to plasmid-containing B. subtilis inocula were in the ratios 1:1, 1:0.1, 1:0.01, 1:0.001. Different dilutions of the cultures were plated on solidified NB-2 medium (for determination of total cell counts) as well as on NB-2 medium containing erythromycin (5 µg/ml) and lincomycin (50 µg/ml) (for determination of plasmid-bearing cell counts).
479
development of competence, aliquots were removed from the growing culture each hour, starting immediately after addition of the inoculum. pMG36enpr was added followed by incubation for 60 min at 37 °C with vigorous shaking. Thereafter, the mixture was plated onto NB-2Em agar plates. Plates were incubated overnight at 37 °C. A low number of competent cells in the growing culture was first detected after 10 hours. The number increased during the next 3 to 4 hours in the stationary phase and then started to decrease until termination of the experiment after 16 hours (Fig. 1). Highest transformation rates observed were ca. 6 × 103/µg DNA. When B. subtilis was grown in chocolate milk a very different kinetic of development of competence was observed. First we noted that growth rate as well as cell yield were higher than in MC(-)-medium, indicating that chocolate milk was apparently a complex and rather good growth medium for B. subtilis. It was, however, a suitable medium for development of competence. This had already been observed by Bräutigam et al. [3] and Zenz et al. [11]. Competent cells, the presence of which was again tested for by removal of aliquots from the growing culture and addition of pMG36enpr, were identified throughout all stages of growth (Fig. 2a). Fluctuations in the number of competent cells were observed with three peaks immediately after inoculation, at the end of exponential growth, and after several hours into stationary phase, respectively. For the peak immediately after inoculation carry-over of competent cells from the stationary over-night culture seems to be most likely. The highest transformation rate detected was 6 × 103/µg DNA after ca. 6 h into stationary phase. It should be mentioned that – although some of the absolute numbers of colony forming units on the agar plates were rather low -– comparable results were obtained in all four repetitions of the experiment. In order to apply a more realistic model concerning the uptake of DNA from food stuff by B. subtilis, we added pMG36enpr to the culture once together with the B. sub-
Quantitation of pMG36enpr DNA by polymerase chain reaction (PCR) During the transformation experiment samples were taken every hour, frozen, and later analyzed for the amount of pMG36enpr DNA. Different fragments of the 6.0 kb pMG36enpr plasmid were amplified by PCR from different dilutions of the plasmid as described by Lick et al. [7], to determine the detection limit for each fragment. For the amplification of a 4.2 kb fragment [7], the detection limit was 0.1 µg/ml in chocolate milk.
Results and Discussion Bacillus subtilis growing in batch culture has been described to develop competence during transition from exponential to stationary growth phase only. As one prerequisite for development of competence, growth in minimal salts instead of complex medium has been identified [4, 6]. We could confirm these results by growing B. subtilis in MC(–)-medium [11]. Exponential growth was observed between 2 and 10 h after inoculation. To test for
Fig. 1. Development of competence of B. subtilis in MC(–)medium. Solid symbols: viable counts (cfu/ml); open symbols: transformed cells counts (cfu/µg DNA).
480
A. Wittke et al.
tilis inoculum. In this case aliquots were removed from the culture and immediately plated onto NB-2-Em agar plates. One hour after inoculation the number of EmR cells was almost exactly the same as in the experiment where aliquots had been withdrawn (Fig. 2b). However, in contrast to our expectations, the number of EmR cells did not increase in parallel to the number of total cells of the culture. Instead, the number of EmR cells decreased within the next two hours, and then slightly increased until the end of the experiment at 24 h after inoculation. The final number of EmR cells at 24 h was ca. 8 × 103/µg DNA. If growth parallel to the total number of cells of the culture would have occurred, the number would have reached ca. 106/µg after 24 h. One interpretation could be that plasmid-bearing cells being present in low amounts
in a culture of otherwise plasmid-free cells suffer from suppression by the plasmid-free cells under non-selective conditions. To test for the latter possibility, different proportions of plasmid-bearing and plasmid-free cells were inoculated from overnight cultures together into chocolate milk at 37 °C. The experiment was carried out twice with almost identical outcome. Growth was monitored by plating different dilutions of the cultures on agar plates with Em (for plasmid-bearing cells) or without Em (for total cells) (Fig. 3a–e). From the samples taken immediately after inoculation the following proportions (total cells:plasmidbearing cells) were calculated: 1:0.38 (Fig. 3a), 1:0.027 (Fig. 3b), 1:0.002 (Fig. 3c), 1:0.00014 (Fig. 3d), 1:1 (Fig. 3e; control 1, plasmid-bearing cells only), 1:0 (Fig. 3f;
Fig. 2. Development of competence of B. subtilis in chocolate milk. a: DNA added to aliquots; b: DNA added once together with the inoculum; solid symbols: viable counts(cfu/ml); open symbols: transformed cells counts (cfu/µg DNA).
Fig. 3. Growth of pMG36enprharbouring and plasmid-free B. subtilis cells, inoculated in varying proportions (plasmid-free cells:plasmid-bearing cells) from overnight cultures. a: 1:0.38; b: 1:0.027; c: 1:0.002; d: 1:0.00014; e: plasmid-bearing cells alone; f: plasmid-free cells alone. Solid symbols: total cells; open symbols: plasmid-bearing cells.
Transformation of Bacillus subtilis in Chocolate Milk
control 2, plasmid-free cells only). It should be noted, that in all cases growth in chocolate milk was in the absence of Em. Several observations were made. i) Growth of plasmid-bearing cells was always slower than corresponding growth of the total cultures. The increase of cell numbers of plasmid-bearing cells within the first four hours was by factors between ca. 25 and 39. This was also true for the culture consisting of plasmid-bearing cells only (Fig. 3e). In contrast, total numbers of cells increased within the first four hours by factors of ca. 150 to 400, with 400 being the increase for the culture consisting of plasmid-free cells only. ii) The increase in plasmidbearing cell numbers in mixed cultures (Fig. 3a–d) was highest for the culture with the highest proportion of plasmid-bearing cells (Fig. 3a). The increase during the first 16 h in the latter case was by a factor of 4.6 × 103, while it was between 6.4 and 8.5 × 102 during the first 12 (Fig. 3b) or 16 h (Fig. 3c and d) for the other three cultures. The slightly delayed growth of plasmid-bearing cells in relation to total (mostly plasmid-free) cells as described under i) may partly be responsible for the reduced multiplication of the plasmid-bearing cells during the first 4 h, since the growth rate of the cells within a culture is determined by the growth rate of the majority of cells. Thus, the plasmid-bearing cells in a mixed culture were not able to achieve their highest growth rate between 4 and 8 h, since the culture as a whole was already beyond is maximal growth rate. iii) The highest absolute cell numbers reached at one point during the 48 h measuring period were comparable for all six cultures, being between 3 and 8 × 109 per ml. Theses numbers were reached, however, at different incubation times: 16 h for the cultures shown in Fig. 3a and 3e, 24 h for those shown in Fig. 3b, 3c, an 3d, and 48 h for growth of plasmid-free cells as shown in Fig. 3f. iv) As a consequence of the observations in iii), all cultures except the one consisting of plasmid-free cells only showed a decline of the absolute cell numbers during stationary phase. The decline was highest in the culture consisting of plasmid-bearing cells only, resulting in a reduction by more than two logs. In all the mixed cultures the number of plasmid-bearing cells declined by less than two logs. It should be noted that in none of the mixed cultures the reduction of the number of plasmid-bearing cells would be sufficient to account for the reduction of the total cell numbers. v) Growth of plasmid-bearing cells in chocolate milk under non-selective conditions did not result in a detectable loss of the plasmid (Fig. 3e). The differences in cell numbers obtained from colony counts on plates with or without Em fell within the regular variations (± 30%) observed for this technique. On the basis of the results obtained with the mixed cultures, the experiment shown in Fig. 2b may be interpreted as follows. i) Uptake of pMG36enpr DNA into the cells appears to take place during the first hour only. Thereafter, the DNA in the medium may become inaccessible for transformation by association with material present in chocolate milk, such material being e.g. casein. Degradation can be excluded for the first 6 h, since the DNA was detectable by PCR at a level of 0.1 µg/ml dur-
481
ing this period (data not shown). ii) Lack of growth of the EmR cells detectable after 1 h can neither be explained by the suppressing activities of the plasmid-free cells, which would only results in a decreased growth rate but not in total inhibition of growth, nor by the lag-phase observed for growth of plasmid-bearing in contrast to plasmid-free cells. Only if this lag-phase was at least three to four hours after uptake of pMG36enpr, the growth curve of the EmR cells could be explained. iii) We propose that the kinetics of appearance of EmR cells is the consequence of two different kinetics: the first being represented by the loss of pMG36enpr under non-selective conditions from cells shortly after uptake of the plasmid, and the second being represented by the growth of a small fraction of EmR cells which have established themselves as permanently pMG36enpr-transformed cells under non-selective conditions. This would mean that of the ca. 5 × 102 EmR cells present after 1 h only about 5 would become permanently transformed and would be responsible for the growth curve observed after the fourth hour after inoculation.
Conclusions One of the hazards described for genetic engineering of bacteria is the transfer of the recombinant DNA into other bacteria of the environment. Especially if recombinant plasmids are used the question will not be whether transfer does occur but to what extend it does occur. The results of this communication allow some quantitative calculations to be made for our model system. When 0.1 µg recDNA and 106 B. subtilis cells are present in one ml of chocolate milk, about one of each 2000 cells will take up DNA. Of these, one of each 100 cells will become permanently transformed. Upon growth, these cells will be lagging by about one log behind the plasmid-free cells. In this model, two numbers have to be questioned. The amount of 0.1 µg recDNA, if considered to be introduced into the chocolate milk e.g. via an additive isolated from a genetically engineered microorganism, is very high. Treatment with nuclease would easily lower the amount to less than 0.1 ng/ml. This would lower the number of transformants concomitantly. The number of 106 B. subtilis cells per ml is also very high. Milk, contaminated with such an amount of saprophytic cells, would already be considered “spoilt”, it would no longer be acceptable for consumption. A contamination e.g. at the process of aseptic filling would result in a likely contamination with only a few cells per milk container with a volume of 1 l. Extrapolating our transformation numbers to a chocolate milk containing 0.1 ng recDNA and 10 B. subtilis cells per ml, 108 ml – corresponding to 100 tons of chocolate milk – would be needed to give rise to 5 permanently transformed cells. Needless to say that even this low number of transformation events is still overestimated i) because of the linear extrapolation, ii) because of the still rather high number assumed for contaminating B. subtilis cells, and iii) because of the assumption that pure plasmid DNA was introduced via an additive. Under realistic con-
482
A. Wittke et al.
ditions, the plasmid would only constitute a minor fraction of the assumed amount of 0.1 ng DNA present per ml.. The majority of the non-plasmid, non-recDNA could act as a competitor for plasmid-uptake, reducing the chances for plasmid transformation even further. Acknowledgements We like to express our thanks to Katrin Schneede and Meike Groszek for excellent technical assistance. This work was in part supported by the German Federal Ministry for Education, Science, Research, and Technology (Project No. 0311049 and 0311049A).
6. 7.
8. 9.
References 1. Ashikaga, S., Nanamiya, H., Ohashi, Y., Kawamura, F.: Natural genetic competence in Bacillus subtilis natto OK2. J. Bacteriol. 182, 2411–2415 (2000). 2. Bauer, F., Hertel, C., and Hammes, W. P.: Transformation of Escherichia coli in foodstuffs. Syst. Appl. Microbiol. 22, 161–168 (1999). 3. Bräutigam, M., Hertel, C., and Hammes, W. P.: Evidence for natural transformation of Bacillus subtilis in foodstuffs. FEMS Microbiol. Lett. 155, 93–98 (1997). 4. Dubnau, D. and Roggiani, M.: Growth medium-independent genetic competence mutants of Bacillus subtilis. J. Bacteriol. 172, 4048–4055 (1990). 5. Errington, J. and Mandelstam J.: Use of a lacZ gene fusion to determine the dependence pattern of sporulation operon
10.
11.
spoIIA in spo mutants of Bacillus subtilis. J. Gen. Microbiol. 132, 2967–2976 (1986). Hauser, P. M. and Karamata, D.: A rapid and simple method for Bacillus subtilis transformation on solid media. Microbiology 140, 1613–1617 (1994). Lick, S., Schneede, K., Wittke, A., Heller, K. J.: Detection without sample processing of yoghurt starters and DNA in milk products by PCR. Kieler Milchwirtschaftl. Forsch.ber. 51: 15–25 (1999). Lorenz, M. G. and Wackernagel, W.: Bacterial gene transfer by natural genetic transformation in the environment. Microbiol. Rev. 58, 563–602 (1994). van de Guchte, M., Kodde, J., van der Vossen, J. M., Kok, J., and Venema, G.: Heterologous gene expression in Lactococcus lactis subsp. lactis: synthesis, secretion, and processing of the Bacillus subtilis neutral protease. Appl. Environ. Microbiol. 56, 2606–2611 (1990). von Wright, A. and Sibakov, M.: Genetic Modification of Lactic Acid Bacteria. In „Lactic Acid Bacteria, Microbiology and Functional Aspects“ (S. Salminen and A. von Wright, Eds.), pp. 161–210. Marcel Dekker Inc. New York (1998). Zenz, K. I., Neve, H., Geis, A., and Heller, K. J.: Bacillus subtilis develops competence for uptake of plasmid DNA when growing in milk products. Syst. Appl. Microbiol. 21, 28–32. (1998).
Corresponding author: Knut J. Heller, Institute for Microbiology, Federal Dairy Research Center, Hermann-Weigmann-Str.1, 24103 Kiel, Germany Tel.: ++49-(0)431-609-2340; Fax: ++49-(0)431-609-2306; e-mail: [email protected]