A major goal of our study was to develop a protocol for efficient natural SAG transformation and utilize competence genes we found in clinical SAG genomes. However, for laboratory tests, we wanted to compare natural transformation efficiency with other methods of introducing DNA into SAG cells. In addition, development of efficient transformation scheme would be beneficial for not naturally competent clinical SAG strain. Electrotransformation was successfully adopted for many Gram-positive bacteria such as Lactococcus lactis (Holo and Nes 1989) and S. pyogenes (Sitkiewicz and Musser 2006); therefore, we started from standardization of electrotransformation protocol for SAG.
To test if the DNA can be efficiently introduced to SAG cells by electrotransformation, we first prepared electrocompetent cells using modified protocols used for other streptococci (Holo and Nes 1989; Dunny et al. 1991; Smith et al. 1995; Sitkiewicz and Musser 2006). On the contrary to published methods, we used BHI rich medium for SAG cultivation with the addition of threonine and glycine. Simultaneous medium supplementation by threonine and glycine was intended to influence on cell wall synthesis. The initial test medium consisted of BHI with 0.5 M sucrose, 40 mM DL-threonine, and 0.1 M glycine. Bacteria were grown without aeration until the culture reached OD600 ca. 0.3, indicating the early exponential phase of growth. The culture was chilled on the ice for 30 min and cells were collected by centrifugation at 4 °C. The supernatant was removed, the bacterial pellet was washed twice using cold 0.5 M sucrose, and finally, bacteria were re-suspended in 0.5 M sucrose with 15% glycerol. The bacterial suspension was divided into single-use portions and frozen at − 80 °C or used directly for electrotransformation. Freshly prepared or thawed on the ice 50 μl portions of competent cells were added to cold electroporation 0.1 cm cuvettes. Then ca. 200 ng of control plasmid pGKV2 was added to the cuvette. Bacteria with DNA were incubated on the ice for 30 min. Next, cuvette was inserted into a sample chamber in Gene Pulser Bio-Rad system and electric pulse was applied with the capacitance of 25 μF, the voltage of 2.5 kV, and resistance of 200 Ω. Warm medium with sucrose (without antibiotics) was added to bacteria immediately after the pulse. Bacteria were incubated for about 3 h and then plated onto the selective medium with erythromycin. Transformation efficiency was in the range of 102 per 1 μg plasmid DNA.
Despite the fact that the cells prepared using this method exhibited low competence, the bacterial culture grew poorly in the test medium. Therefore, we analyzed which of the medium components is crucial for growth and transformation efficiency. We checked several variants of the media: supplemented or not with glycine, with different concentration of sucrose, and with the presence of L-, D-, or DL-isomers of threonine and incubation in presence of an increased concentration of CO2 during growth in the supplemented medium.
Interestingly, we observed that growth of test 4194/05 strain in the supplemented medium was reduced by at least 50% compared to growth in medium without any additional substance. Elimination of glycine from the medium supplemented by sucrose and threonine significantly improved 4194/05 strain growth rate. However, no transformants were observed on selective medium after multiple independent attempts of electrotransformation of competent cells prepared without glycine and using control plasmid pGKV2. Therefore, the addition of glycine seems to be crucial for preparation of SAG electrocompetent cells. Moreover, we noticed that elimination of sucrose with the simultaneous presence of threonine and glycine positively influenced on the growth of 4194/05 strain. Cells cultivated in such composed medium gave on average of 24 transformants/μg of plasmid DNA. Then, we tested medium supplemented by a lower concentration of sucrose (0.25 M, 40 mM DL-threonine, 0.1 M glycine) since sucrose could protect cells with defective cell wall caused by treating by glycine and threonine. In result, we observed transformation efficiency in the range of 104 per 1 μg plasmid DNA. Next, we tested the influence of threonine isoform on the transformation efficiency. D-threonine supplementation seemed to worsen transformation level. Addition of DL- or L-threonine to the medium resulted in the same transformation efficiency, ca. 2 × 104 transformants/μg of plasmid DNA. We also tested an influence of the electric pulse parameters on the transformation efficiency. Higher resistance used during the electroporation decreased slightly the average transformation efficiency: 200 Ω, 2.3 × 103; 400 Ω, 1.5 × 103; and 600 Ω, 5 × 102 transformants/μg plasmid DNA. Incubation in presence of 5% CO2 was crucial for the growth of 4194/05 strain in restricted conditions caused by sucrose, glycine, and threonine. In conclusion, the highest transformation efficiency obtained for 4194/05 strain was therefore around 104 transformants per 1 μg of pGKV2 plasmid DNA. Summary of transformation efficiencies obtained for 4194/05 are presented in Tables 3 and 4.
Using electrocompetent 4194/05 cells, we were unable to obtain transformants/integrants after attempts to introduce linear DNA 5’L-spc-R3’. It suggested degradation of a linear DNA, a defect in recombination process, or too low efficiency to obtain a mutant with a linear fragment integrated via double crossing-over into the chromosome.
To check if the recombination system of 4194/05 strain works properly, we used pLR plasmid (with 5’L-spc-R3’ cassette), which is unable to replicate in Gram-positive bacteria. Recombination machinery of 4194/05 strain after the uptake of pLR should force into integration the plasmid into the chromosome, most probably via Campbell recombination. As a result of the first transformation trial using previously prepared electrocompetent cells, characterized by plasmid transformation efficiency in the range 104, several transformants grew on selective medium with spectinomycin. The downstream analysis showed that the whole vector was incorporated into the chromosome of 4194/05 strain, resulting 4194/05_Camp. Integration of pLR into the chromosome was verified by PCR using specific primers (upLF and downRR). This result indicated that 4194/05 strain is able to integrate foreign DNA into the chromosome. The most probable explanation of inability to uncover colonies on selective plates after transformation with linear DNA is the low-efficiency transformation of 4194/05 compared with the frequency of double crossing-over. It seems that double recombination in the genetic background of 4194/05 strain occurs less frequently than the frequency of plasmid introduction into 4194/05 competent cells.
We analyzed the ability of another SAG strain to efficient transformation by media modification. We verified the best-tested conditions for electroporation procedure in case of another SAG strain, 980/01. Briefly, the strain was cultured in BHI with the presence of 40 mM DL-threonine, 0.1 M glycine, and 0.25 M sucrose, washed by 0.5 M sucrose, and treated by electric pulse (25 μF, 2.5 kV, 200 Ω, 0.1 cm cuvettes). Using control plasmid pGKV2, the average electroporation efficiency was 9 × 103/μg DNA. The reveled efficiency of 980/01 strain was slightly lower than the efficiency of 4194/05 strain (2 × 104/μg plasmid DNA). This result showed also that using tested protocol competent, SAG cells were obtained with the similar effectiveness. Although the transformation ability of both strains was comparable, we additionally examined if 980/01 strain could be electrotransformed with linear DNA. The electroporation procedure was repeated using the 5’L-spc-R3’ cassette amplified on the pLR plasmid template. Transformants were selected on medium with spectinomycin. After incubation, about 100 spectinomycin resistant colonies grew onto selective medium. Further analysis, using primer pairs upLF and spcR, or spcF and downRR, for spc cassette and for flanks of deleted codY gene, verified replacement of the codY gene by spectinomycin resistance cassette in 980/01 strain in ca. 50% of transformants.
Interestingly, spectinomycin resistant colonies grew on the medium with spectinomycin concentration 2× MIC (minimal inhibitory concentration) for the tested strains without introduced spc cassette. Spectinomycin resistance is based on blocking of protein translation and the antibiotic interacts with a small subunit of the bacterial ribosome. Presence of spontaneous spectinomycin resistance might be explained by a simple mutation of the gene encoding a ribosomal protein which interacts with the antibiotic. On the other hand, SAG may encode an unknown enzyme which modifies the antibiotic or may encode not-specific efflux pump, which provides antibiotic molecules transport out of the cell.
Transformation of protoplast
Since transformation efficiency of 4194/05 electrocompetent cells was low, we decided to test alternative approach of protoplast transformation. Thus, an overnight culture of 4194/05 strain in BHI was collected by centrifugation (10 min, 3000 rcf, at room temperature). Then, bacteria were re-suspend in 5 ml of protoplastization buffer (BHI supplemented with SMM buffer (0.5 M sucrose, 0.02 M maleate, 0.02 MgCl2)) (Woskow and Kondo 1987) or saline solution. Sample from saline solution was diluted and plated onto BHI medium to count bacteria. Cells in protoplastization buffer were incubated with gentle shaking at 37 °C. Additionally, cells were treated by enzymes removing cell wall, lysozyme (10 mg/ml, Sigma), and mutanolysin (25 U/ml, Sigma). To prevent protoplasts from the rupture after cell wall digestion, they were stored into isotonic solution, such as BHI medium supplemented with 0.5 M sucrose, 0.02 M maleate, and 0.02 M Mg2+ ions.
Initially, we examined if protoplasts were generated in tested conditions and if they could regenerate after plating onto the solid medium. Bacteria re-suspended in protoplastization buffer were incubated at 37 °C with gentle mixing. After 30 min, 2 or 4 h of incubation samples were collected. Each sample was diluted from 10−5 to 10−10 in H2O and BHI supplemented by SMM. Then, dilutions were plated onto two variants of medium BHI (standard broth) or BHI with 0.5 M sucrose (regeneration broth). A number of bacteria diluted in BHI supplemented by SMM and grown onto BHI with sucrose indicated the number of all bacteria. A number of bacteria diluted in H2O and grown onto BHI indicated the number of bacteria that were not in protoplast stage. Protoplastization of the cells was verified by microscope observation.
Three independent trials were performed. On average, after 30 min, cell wall digestion of 4194/05 strain ca. 67% (15 × 1013 vs 5 × 1013) of cells were cell wall free (cfu on regeneration broth vs cfu on standard broth). Two hours of incubation gave around 99.6% cell wall free cells (15 × 1012 vs 6 × 1010). A similar result, of 99.9% of cell wall free cells, was obtained after 4 h of incubation. However, the number of all cells dramatically decreased (6 × 1010 vs 2 × 107) after 4-h incubation in protoplastization buffer. Therefore, in further tests of protocol development, the incubation was carried out for 2 h. For electroporation, after 2-h incubation, protoplasts were washed in 0.5 M sodium succinate and centrifuged for 10 min at 3000 rcf and re-suspended in sodium succinate. Then protoplasts were transferred to cold 0.2-cm gap cuvette. Approximately 1 μg of control plasmid pGKV2 or chromosomal DNA of 4194/05_Camp strain (with marker spc cassette) was added. After 30 min on ice, electroporation was performed. Initial parameters were capacitance of 25 μF, the resistance of 400 Ω, and voltage of 0.5 kV. Subsequent attempts tested voltage of 0.2, 1, and 2 kV and resistance of 200 and 600 Ω. BHI medium supplemented by SMM was added to bacteria immediately after the electric pulse. Bacteria were incubated for 6 h in growth medium supplemented by 0.25 M sucrose to allow cell wall regeneration. Then, they were plated onto selective medium.
We did not observe regeneration after the shock. After ten trials with different electric conditions, the number of cfu grown onto regeneration broth was comparable to the number of cfu onto standard broth (on average 4.6 × 107 vs 6.6 × 107). Although the low strength of electric pulse was used, it could damage protoplasts. Whereas, colonies grown onto broth originated from the part of not-digested cells. The second reason might be that the regeneration broth was good enough for regeneration protoplasts after cell wall digestion but not for recovering them after an electric shock.
In conclusion, we were unable to setup electrotransformation parameters that allowed 4194/05 strain to survive the electric pulse. Protoplasts not subjected to the electric pulse were able to regenerate and form colonies, but cells after electric pulse were not. We think that SAG cells require very gentle parameters of the electric pulse to be able to regenerate afterward. In our hands, the protoplast electrotransformation does not work as a way to introduce DNA into SAG cells.
Rather unsuccessful attempts with efficient electrotransformation of SAG cells or protoplasts led us to tests of the efficiency of natural transformation. Replacement of the target gene by resistance marker by the double recombination was observed in other SAG strains (Bauer et al. 2018), as a result of the natural transformation. Putative natural competent SAG strains could be selected based on the detection of com operons in their genomes. Analysis of 4194/05 genome revealed the presence of comAB and comCDE operons. Therefore, putative natural CSP (product of comC gene) should be present in the 4194/05 culture medium. An appropriate amount of CSP in culture medium should induce competent state based on quorum sensing principles. So, the addition of overnight culture medium to fresh bacterial culture might increase the concentration of CSP. We attempted to induce natural competence of 4194/05 strain without and with using the post-culture medium. The post-culture medium was sterilized neither through autoclaving nor through filtration as temperature could inactivate the peptide, and membrane could bind the peptide. Briefly, an overnight culture of 4194/05 strain was diluted 1:10 in fresh BHI and incubated for 70 min at 37 °C. Then, chromosomal DNA of 4194/05_Camp (200–400 ng/ml), either with or without a post-culture medium (1:10), was added. After 3-h incubation, bacteria were plated onto the selective medium with spectinomycin. We did not observe any recombinants of 4194/05 strain after three independent trials of bacteria incubation with DNA, neither with nor without the addition of post-culture medium.
Next attempt of enforcing competence state of 4194/05 strain was based on recent reports. As it was observed, the sub-lethal concentration of antibiotics activates stress response and expression of genes located near the oriC including competence genes (Prudhomme et al. 2006; Slager et al. 2014). For competence stimulation of 4194/05 strain, we used 25 μg/ml kanamycin and 10 μg/ml streptomycin. Like previously, diluted bacteria culture was incubated 70 min, then chromosomal DNA of 4194/05_Camp and antibiotics were added and further incubation was carried out. After 3 h bacteria were plated onto selective broth. Medium supplementation by a sub-lethal concentration of antibiotics was not successful as well, no recombinants of 4194/05 strain grew onto selective broth. This failure could be explained by that the concentration of antibiotics did not provide proper stress condition and DNA uptake system was not induced.
In the last approach towards induction of 4194/05 strain competence, we used artificial CSP. Peptide sequence DSRIRMGFDFSKLFGK (Lipopharm) was based on genomic sequence of 4194/05 strain and matched the sequence used and published by Håvarstein (Håvarstein et al. 1997). Firstly, refreshed overnight culture (1:10) was incubated for 2 h at 37 °C. Next natural competence was induced by addition, to the bacterial culture, CSP at the concentration 5 ng/ml, 50 ng/ml or 0.5 μg/ml together either with chromosomal DNA of 4194/05_Camp (ca. 3 μg/ml) or with the linear PCR product of 5’L-spc-R3’ cassette (200–400 ng/ml). Then, bacteria were incubated for 3 h at 37 °C and plated onto selective medium. No positive results were obtained from two independent transformation tests. We hypothesized that either competence system of this particular strain did not work properly or synthetic peptide was inactive.
To check if the synthetic CSP was functional, we tested its activity on 17 other SAG strains from our collection (Obszanska et al. 2016). Two-hour cultures were divided into three separate samples. Each was supplemented by (i) CSP and chromosomal DNA of 4194/05_Camp, (ii) only chromosomal DNA, or (iii) only CSP. The final concentration of CSP was 100 ng/ml and chromosomal DNA ca. 3 μg/ml. After 3-h incubation, samples were plated onto selective medium. In the first variant (CSP and DNA), we obtained recombinants in case of seven strains (980/01, 2160/05, 4020/05, 2088/07, 2721/07, 178/09, 1365/09). We checked and verified the presence of comDE genes in genomes of these seven strains using specific primers. Since the unpublished genomic sequences of few of those strains were available for our laboratory (980/01, 4020/05, 2088/07, 178/09, 1365/09), we could verify also a sequence of comC present in particular strains. The identical sequence to synthetic CSP was observed in genomes of 980/01, 4020/05, 178/09, and 1365/09 strains. Interestingly, none known homolog of comC was detected in the genome sequence of 2088/07 strain. Transformation efficiency for strains was in the range of 1 recombinant to 100 recombinants per 1 μg of chromosomal DNA. Four (4020/05, 2721/07, 178/09, 1365/09) of seven strains that gave positive results in the first variant of the experiment yielded also colonies after incubation with DNA alone, without the addition of synthetic CPS (Table 5). However, the number of spectinomycin colonies was at least twofold lower than in the first variant with synthetic CSP. These results indicated that the transformation efficiency using synthetic CSP is strain dependent. We verified the presence of spectinomycin resistance cassette in PCR using specific primers and chromosomal DNA of recombinants as templates. In addition to the transformation using chromosomal DNA, we tested transformation and integration of the linear PCR product. We decided to use strains that were shown to be the most efficient in tests with chromosomal DNA (4020/05 and 1365/09). The workflow for the transformation with linear 5’L-spc-R3’ cassette (200–400 ng/ml) was the same as in transformation with chromosomal DNA. As a result of experiment, three spectinomycin, resistant colonies of 4020/05 strain, and two of 1365/09 strain were obtained The presence of spc cassette after transformation was verified by PCR.