The effect of multiple
consecutive transfers of a
single plasmid in E. coli
Siem de Haan *, under supervision of Tania Darphorn (MSc) and Benno ter Kuile (dr.) 10-05-2020
Dept. of Molecular Biology & Microbial Food Safety, University of Amsterdam, Swammerdam Institute of Life Sciences, Amsterdam, The Netherlands
* s.j.dehaan@amsterdamumc.nl
SUMMARY
Before the current COVID-19 pandemic, the World Health Organization already highlighted antibiotics resistance as an increasing threat associated with high global mortality rates in the coming decades. Besides the use of antibiotics in human medicine, the high consumption of antibiotics in livestock plays a major role in the development of antibiotics resistance. Conjugation of plasmids is the most prominent mechanism by which bacteria can acquire resistance against antibiotics. Although it has been demonstrated that plasmids can be transmitted through the human food chain, multiple transfers of a single plasmid between bacteria have not been well investigated. Therefore, we investigated the effect of 12 consecutive transfers of a single plasmid on the transfer efficiency and antibiotics resistance of E. coli, and on the DNA sequence of the plasmid itself. Results indicated that multiple plasmid transfers have no effect on the transfer efficiency and antibiotics resistance of E. coli, nor on the sequence of well-annotated genes on the plasmid, except for the gene pilV. The
observation that transfer efficiency, antibiotics resistance and plasmid integrity remain stable after 12 transfers underlines the robustness of the conjugation process as a driver of antibiotics resistance.
INTRODUCTION
Before the current COVID-19 pandemic, the World Health Organization already highlighted antibiotics resistance as an increasing threat associated with high global mortality rates in the coming decades (WHO, 2019). Not only the use of antibiotics in human medicine, but also the high consumption of antibiotics in livestock plays a major role in the development of antibiotics resistance (Landers et al., 2012). China has been estimated to be the leading country in the agricultural use of antibiotics, consuming 30% of the global antibiotics production by 2030 (van Boeckel et al., 2015).
There are several ways in which the extensive use of antibiotics in both livestock and human medicine can increase antibiotics resistance. One of the main mechanisms by which bacteria acquire resistance to antibiotics is horizontal gene transfer (Mazel & Davies, 1999). Horizontal gene transfer can further be divided into transformation, transduction and conjugation (Thomas & Nielsen, 2005), of which conjugation is the most prominent mechanism (Nazarian et al., 2018). Conjugation is a process in which a donor bacterium transfers a plasmid to an acceptor bacterium (Norman et al., 2009). Plasmids are circular DNA structures that encode so-called ‘backbone’ genes and ‘accessory’ genes
(Harrison & Brockhurst, 2012). Backbone genes are associated with primary functions such as replication and the formation of a conjugation pilus, whereas accessory genes yield potentially beneficial traits for the bacterium, including antibiotics resistance (Millan, 2018).
In agriculture, the extensive use of antibiotics exerts a selective pressure on bacteria that carry plasmids associated with antibiotics resistance, also called R-plasmids (You & Silbergeld, 2014). Since these bacteria are incorporated in the human food chain, they can transfer their R-plasmids to human pathogenic bacteria in the gastrointestinal tract (Levy et al., 1976). Indeed, similar plasmid sequences containing extended-spectrum ß-lactamase (ESBL) genes have been found in E. coli from poultry and humans (Leverstein-van Hall, 2011). Altogether, the transmission of R-plasmids from agricultural to human pathogenic bacteria can result in human disease (Händel et al., 2015).
Although it has been demonstrated that plasmids can be transmitted through the food chain, the effects of multiple transfers of a single plasmid have not been well investigated. In fact, it is still unknown if any genotypic changes occur in the plasmid, or any phenotypic changes in the acceptor bacterium after a plasmid is transferred multiple times between bacteria. Therefore, the aim of this research is to investigate the effect of 12 plasmid transfers on the transfer efficiency and antibiotics resistance of E. coli, and on the integrity of the plasmid sequence itself. Transfer efficiency will be measured by counting Colony Forming Units (CFU) on agar plates, antibiotics resistance by Minimum Inhibitory Concentration experiments and plasmid integrity by plasmid sequencing analysis.
MATERIALS AND METHODS
Bacterial strainsThe E. coli strains that were used for the transfer experiments were the Keio indole knockout JW3686 (provided by DharmaconTM) and the MG1655 YFP (provided by MB Elowitz). The JW3686 strain is known of chromosomal resistance against kanamycin, whereas the MG1655 strain is characterized by chromosomal resistance against chloramphenicol. The MG1655 YFP was also carrying an ESBL plasmid (3170.1, provided by the Netherlands Food and Consumer Product Safety Authority).
Growth medium, agar plates and antibiotics
All bacteria were grown in Evans medium containing glucose (55 mM, pH 6,9), which was prepared following the same protocol as Händel et al. (2015). Bacteria were always grown and starved in a 37 °C 200 RPM shaking incubator. During the starvation part of the transfers, the same Evans medium was used but without glucose. After the transfer experiments, bacterial suspensions were poured on LB agar plates containing 64 µg/ml of either ampicillin, chloramphenicol, kanamycin, or a combination of these. All antibiotics stocks, including tetracycline, were 0.2-µm filter-sterilized and diluted to a concentration of 10 mg/mL, then stored at 4°C and disposed after 14 days (Händel et al., 2015).
Plasmid transfer experiments + storage
For the first transfer, three samples of the donor (MG1655) and one sample of the acceptor (JW3686) were grown overnight in 10 mL Evans medium (55 mM glucose), containing 32 µl ampicillin for the donor (MG1655) and 32 µl kanamycin for the acceptor (JW3686). For some even number transfers, 16 µl of chloramphenicol was used when the MG1655 acceptor showed no growth after overnight incubation. After overnight incubation, samples were centrifuged (15 minutes, 4.4K RPM) and cell pellet was diluted in 5 mL Evans medium (no glucose) for the 4h starvation.
Following this starvation, OD was measured at 600 nm and the volume of bacterial suspension needed for an OD600 of 0.25 in 10 mL Evans medium (no glucose) was calculated. For each donor
sample, this specific volume of donor and acceptor was added to an Erlenmeyer flask containing 10 mL Evans medium (no glucose). For each donor sample, this step was performed twice to make an 1h and 24h incubation sample. Thereafter, the six Erlenmeyers were placed in the incubator.
After 1h or 24h of incubation, samples were taken from the incubator and diluted in sterile MilliQ (dilutions of 10-0 – 10-5, first dilution 100 µl sample in 900 µl MilliQ). Next, 50 µl of the 10-0, 10-1 and 10-2 dilution of each sample was pipetted on agar plates containing ampicillin and kanamycin. Furthermore, 50 µl of the 10-3, 10-4 and 10-5 dilutions of each sample was pipetted on agar plates containing ampicillin, as well as on agar plates containing kanamycin. Plates were incubated overnight at 37 °C. On the next day, the number of Colony Forming Units (CFU) on the plates was counted by eye. For each sample, 10 mL Evans medium with glucose was then inoculated with a single colony from the kanamycin/ampicillin 1h incubation plate (a transconjugant) and grown overnight for the second transfer. For each sample, one
transconjugant plate was stored at 4°C. The exact same procedure was used for the second transfer. However, since the JW3686 now became the donor, a MG1655 strain without the ESBL plasmid was used as the acceptor. Although the Evans medium of the donor again contained ampicillin, the medium of the acceptor now contained chloramphenicol. After the starvation, 1h or 24h incubation and dilution steps, 50 µl of the 10-0, 10-1 and 10-2 dilution of each sample was now pipetted on agar plates containing ampicillin and chloramphenicol. Furthermore, 50 µl of 10-3, 10-4 and 10-5 dilutions of each sample was pipetted on agar plates containing ampicillin, as well as on agar plates containing kanamycin. For transfer 3, the procedure was exactly the same as for transfer 1, whereas the procedure of transfer 4 was exactly the same as transfer 2. In this way, the plasmid was transferred 12 times from donor to acceptor (Figure 1).
Figure 1: Multiple transfers from donor to acceptor Shown are the first three transfers, starting with the
MG1655 strain as donor, and the JW3686 strain as acceptor. CAM-R: chloramphenicol resistance gene, KAN-R: kanamycin resistance gene.
MIC measurement
The Minimum Inhibitory Concentration of ampicillin, kanamycin, chloramphenicol and tetracycline were determined for the 1h and 24h transconjugants using the Thermo Scientific Multiskan FC (595 nm, 37 °C, shake on). In order to do this, a colony of each stored (4 °C) transconjugant plate was inoculated in 5 mL Evans medium (55 mM glucose) and 32 µl ampicillin. For odd number transfers 32 µl kanamycin was added, and for even number transfers 32 µl chloramphenicol (in some cases 16 µl, when 32 µl inhibited growth entirely).
All samples were grown overnight. The next morning, a 96-well plate was filled with Evans medium (55 mM glucose) and the desired antibiotic concentration to a total volume of 150 µl (Table 1). Thereafter, 5 µl bacteria suspension with a final OD600 of 0.05 in 150 µl were added to all the wells except for the last column. Growth was monitored at 595 nm and displayed using the program SkanIt
(version 6.0.1, ThermoFisher Scientific). An OD595 of 0.2 was defined as the threshold of growth. After filling the 96-well plate, 500 µl of the remaining bacterial suspension was mixed with 500 µl glycerol solution (3.33 M) and stored in a cryotube at -80 °C.
Table 1: layout of the 96-well plate for MIC determination
1 2 3 4 5 6 7 8 9 10 11 12 A (AMP) 2048 1024 512 256 128 64 32 16 8 4 0 No E. coli B (AMP) 2048 1024 512 256 128 64 32 16 8 4 0 No E. coli C (KAN) 2048 1024 512 256 128 64 32 16 8 4 0 No E. coli D (KAN) 2048 1024 512 256 128 64 32 16 8 4 0 No E. coli E (CAM) 2048 1024 512 256 128 64 32 16 8 4 0 No E. coli F (CAM) 2048 1024 512 256 128 64 32 16 8 4 0 No E. coli G (TET) 512 256 128 64 32 16 8 4 2 1 0 No E. coli H (TET) 512 256 128 64 32 16 8 4 2 1 0 No E. coli
Shown are the concentrations of antibiotics in µg/mL. Column 12 contained only 150 µl Evans medium (55 mM glucose).
Isolation and sequencing of the plasmid
Plasmids were isolated from the three 1h samples of transfer 1 and transfer 12 using the QIAGEN Plasmid Maxi Kit (Appendix 1). In addition, samples were cleaned using an ethanol precipitation protocol (Appendix 2) to improve the quality of the sample. Illumina sequencing and long read PacBio sequencing was performed by BaseClear (Leiden, 1993), after which the FASTA files were annotated using RAST (Version 2.0; The SEED Team). The FASTA files were then investigated using PlasmidFinder (Version 2.1; Center for Genomic Epidemiology, 2020) to find similar plasmids in the database, ResFinder (Version 3.2; Center for Genomic Epidemiology, 2020) to reveal antibiotics resistance genes and CLC Workbench (Version 8.1; QIAGEN) to compare gene sequences between the samples of transfer 1 and transfer 12. The sequences of the segments within the PilV gene were assessed in more detail using primer sequences of Turner et al. (2014), Appendix 3.
Data analysis
For the transfer efficiency measurements, a CFU count range of 0-300 was used because some 24h transfers (2,3,4,6,11) yielded a number of transconjugants below the recommended count range of 30-300 (Breed and Dotterer, 1916). For all counts (all plates), CFU/mL values were calculated by multiplying the number of CFU by 20 (20 * 50 µl = 1 mL bacterial suspension) and the dilution factor (10-0 – 10-5). Results were then log
10 transformed, after which the mean log CFU/mL value of the three transconjugant plates of each sample (10-0, 10-1 and 10-2 plate) was calculated.
The mean log CFU/mL value for the donor was calculated by subtracting the mean of the
transconjugant plates from the mean of the ampicillin plates (10-3, 10-4 and 10-5). Likewise, the mean log CFU/mL value for the acceptor was calculated by subtracting the mean of the transconjugant plates from the mean of the kanamycin or chloramphenicol plates (10-3, 10-4 and 10-5), depending on the transfer. Standard deviations of the donor and acceptor were calculated by pooling the variance of the transconjugant plates and the kanamycin or chloramphenicol plates.
Due to small sample sizes and high variation within the ‘technical’ replicates (for instance, the 10-0, 10-1 and 10-2 plates of the transconjugant), no statistical test was performed. Likewise, no statistical test was performed on the MIC results, since the MIC experiments were only meant as control experiments, and therefore not repeated enough to accomplish sufficient sample sizes.
RESULTS
Transconjugants show irregular transfer efficiency pattern over the sequential transfers
In order to investigate the transfer efficiency of each transfer, mean CFU/mL values of the log10 -transformed data of the transconjugants were calculated and are shown in Table 2, including data of the acceptor and donor. For each transfer, the data of the 1h transfers and 24h transfers are shown in
Figure 2 and Figure 3 respectively. The raw untransformed CFU data are included in Appendix 4. Table 2: mean Colony Forming Units of the acceptor, donor and transconjugant.
Shown are the mean (log) ± standard deviation (SD) of the three samples for each transfer after 1h and 24h of incubation.
Transfer Acceptor Incubation Mean CFU/mL acceptor ± SD (log10 values) Mean CFU/mL donor ± SD (log10 values) Mean CFU/mL transconjugant ± SD (log10 values) 1 JW3686 1h 4.21 ± 0.24 4.73 ± 0.41 3.75 ± 0.19 24h 5.77 ± 0.31 6.38 ± 0.47 1.73 ± 0.38 2 MG1655 1h 4.08 ± 0.23 3.95 ± 0.21 4.19 ± 0.26 24h 6.05 ± 0.24 7.00 ± 0.11 0.79 ± 0.06 3 JW3686 1h 6.90 ± 0.97 2.41 ± 0.96 1.37 ± 0.97 24h 5.80 ± 1.05 6.08 ± 1.08 1.97 ± 1.14 4 MG1655 1h 3.83 ± 1.16 6.21 ± 0.32 1.73 ± 0.28 24h 5.34 ±0.45 6.87 ± 0.81 1.35 ± 0.52 5 JW3686 1h 5.56 ± 1.32 5.86 ± 1.48 2.56 ± 1.34 24h 4.62 ± 0.40 4.57 ± 0.31 3.46 ± 0.39 6 MG1655 1h 4.12 ± 1.89 6.31 ± 0.27 1.66 ± 0.45 24h 5.68 ± 0.00 * 6.69 ± 0.54 1.08 ± 0.72 7 JW3686 1h 3.81 ± 0.27 4.13 ± 0.13 4.16 ± 0.27 24h 3.10 ± 0.07 3.14 ± 0.07 4.65 ± 0.04 8 MG1655 1h 4.57 ± 1.26 5.66 ± 0.05 2.03 ± 0.11 24h 1.20 ± 1.54 6.54 ± 0.46 1.28 ± 0.41 9 JW3686 1h 4.62 ± 0.18 4.29 ± 0.24 3.71 ± 0.18 24h 4.26 ± 0.10 4.67 ± 0.03 3.49 ± 0.03 10 MG1655 1h 3.70 ± 0.55 4.97 ± 0.48 3.29 ± 0.55 24h 3.19 ± 1.69 5.67 ± 0.71 2.54 ± 0.77 11 JW3686 1h 3.98 ± 0.41 4.33 ± 0.33 3.72 ± 0.44 24h 6.20 ± 0.47 6.99 ± 0.51 1.12 ± 0.57 12 MG1655 1h 3.56 ± 1.14 4.50 ± 0.95 3.16 ± 0.98 24h 4.59 ± 0.08 5.63 ± 0.01 1.96 ± 0.14
* In this transfer, two samples showed yielded less CFU on the acceptor plates compared to the transconjugant double resistance plates. As a result, subtracting the CFU/mL of the transconjugant plates from the acceptor plates in order to calculate the CFU/mL acceptor yielded a negative CFU/mL value for the acceptor. Therefore, only the mean and SD of the one sample with a positive CFU/mL value for the acceptor is shown.
For the transfers that had incubated for 1 hour, the acceptor showed log CFU/mL values ranging between 3.56 – 6.90, with a mean of 4.41 CFU/mL over all transfers (Figure 2). Especially in transfer 3 and 5, the acceptor yielded strikingly high CFU/mL values (6.90 and 5.56 CFU/mL respectively). Similar to the acceptor, the donor showed a large range of CFU/mL values over all transfers (2.41 – 6.31, mean 4.78 CFU/mL). The donor showed a remarkably low CFU/mL value at transfer 3 (2.41 CFU/mL), whereas transfers 4, 5, 6 and 8 yielded high numbers (6.21, 5.86, 6.31 and 5.66 CFU/mL respectively). Compared to the acceptor and donor, a lower range of CFU/mL values was found for
the transconjugant (1.37 – 4.19, mean 2.94 CFU/mL). CFU/mL values of the transconjugant were especially low at transfers 3-6 and transfer 8 (1.37, 1.73, 2.56, 1.66 and 2.03 CFU/mL respectively). Figure 2: Colony Forming Units of acceptor donor and transconjugant after 1h incubation for each transfer.
Shown is the log CFU/mL + standard deviation for each type at the 12 transfers.
In the same way, results of the transfers with an incubation period of 24 hours are shown in Figure 3. In this case, the acceptor yielded CFU/mL values ranging between 1.20 – 6.21 for all transfers, with a mean of 4.65 CFU/mL. The most remarkable acceptor transfers were transfer 7, 8 and 10, since these transfers yielded low CFU/mL values (3.10, 1.20 and 3.19 CFU/mL respectively). Compared to the acceptor, the donor showed a higher range of CFU/mL values over all transfers (3.14 – 7.00, mean 5.85 CFU/mL). With regard to the donor, the relatively low CFU/mL values of transfer 5, 7 and 9 (4.57, 3.14 and 4.67 CFU/mL respectively) are worth noting. The transconjugant yielded CFU/mL values
ranging between 0.79 – 4.65 CFU/mL with a mean of 2.12 CFU/mL. Interestingly, odd number transfers produced on average a higher CFU/mL than even number transfers (2.74 vs. 1.50 CFU/mL). Figure 3: Colony Forming Units of acceptor donor and transconjugant after 24h incubation for each transfer.
Shown is the log CFU/mL + standard deviation for each type at the 12 transfers.
From the previous two figures, the data of the transconjugant for both the 1h and 24h transfers are combined in Figure 4. For the 1h transfers, the transconjugant results ranged between 1.37 – 4.19 CFU/mL with a mean of 2.74 CFU/mL, compared to a range of 0.79 – 4.65 CFU/mL and a mean of 2.12 CFU/mL for the 24h transfers. Most transfers yielded a higher CFU/mL value for the 1h samples
compared to the 24h samples, except for transfer 3, 5 and 7. The transfer efficiency was not clearly dependent on the number of CFU of the acceptor or donor (Appendix 5).
Figure 4: Colony Forming Units of the transconjugants after 1h and 24h incubation for each transfer. Shown is the mean log CFU/mL + standard deviation for the 1h and 24h samples for each transfer.
With regard to the CFU/mL results shown in Figure 2, 3 and 4, it is important to note that the standard deviations in these figures should not be underestimated by eye, since these are plotted on a log10 axis and are hence compressed compared to the untransformed data (Appendix 4). Moreover, the means of the untransformed donor, acceptor and transconjugant data in Appendix 4 are also accompanied by very high standard deviations. Altogether, the high variance in our dataset made it impossible to meet the assumptions of a proper statistical test.
Nevertheless, it was still possible to recognize rough patterns in these figures. With regard to the previous work of Ivan Fung (Appendix 6), an irregular pattern of transconjugant CFU/mL over the transfers was expected for the 1h samples, whereas a high number of CFU/mL for the odd number transfers compared to the even number transfers was hypothesized for the 24h samples. Based on our findings in Figure 4, the transconjugants after 1h incubation indeed showed an irregular pattern of CFU/mL over the transfers. In addition, Figure 4 also shows the hypothesized CFU pattern for the transconjugants after 24h incubation, with the exception of transfer 11, showing a low number of log CFU/mL (1.12 ± 0.57) where it was expected to be higher (mean log CFU/mL odd number transfer transconjugants was 3.06).
Lastly, three remarkable patterns in Figure 4 are worth noticing. First of all, 1h transconjugant samples seem to have a higher transfer efficiency than the 24h transconjugant samples for most transfers. Secondly, transfers 3, 4 and 6 show a remarkably low transfer efficiency for both the 1h and
24h samples. Third, the 24h samples of transfers 5 and 7 show strikingly higher transfer efficiency compared to the 1h samples, Several biological and methodological explanations for these findings are described in the discussion section.
Minimum Inhibitory Concentration of antibiotics meet expectations
In order to investigate the effect of each plasmid transfer on the antibiotic resistance of the
transconjugants, the Minimum Inhibitory Concentration (MIC) of ampicillin, tetracycline, kanamycin and chloramphenicol were determined for both the 1h and 24h incubation samples of each transfer. Results are shown in Table 3. Due to limited plate reader availability, only the MIC values from transfer 1 to the 1h samples of transfer 6 are based on measurements of 2-3 samples, whereas the MIC results from transfer 6 24h until transfer 12 are based on one sample (SD = 0).
Table 3: mean Minimum Inhibitory Concentration (MIC) values for all the transfers after 1h and 24h incubation. Shown are the mean and standard deviation (SD) of the three samples for each transfer after 1h and 24h of incubation. Transfer Incubation MIC ampicillin
(µg/ml) ± SD MIC tetracycline (µg/ml) ± SD MIC kanamycin (µg/ml) ± SD MIC chloramphenicol (µg/ml) ± SD 1 1h 2048 ± 0 171 ± 60 2048 ± 0 16 ± 0 24h 2048 ± 0 117 ± 15 2048 ± 0 20 ± 9 2 1h 2048 ± 0 64 ± 0 0 ± 0 * 256 ± 0 24h 2048 ± 0 128 ± 0 24 ± 7 427 ± 121 3 1h 2048 ± 0 224 ± 32 2048 ± 0 7 ± 1 24h 2048 ± 0 117 ± 15 2048 ± 0 21 ± 10 4 1h 2048 ± 0 192 ± 52 32 ± 0 427 ± 121 24h 2048 ± 0 171 ± 60 32 ± 0 427 ± 121 5 1h 2048 ± 0 128 ± 0 2048 ± 0 16 ± 7 24h 2048 ± 0 139 ± 40 2048 ± 0 91 ± 38 6 1h 2048 ± 0 149 ± 80 32 ± 0 384 ± 105 24h 2048 ± 0 128 ± 0 32 ± 0 512 ± 0 7 1h 2048 ± 0 96 ± 0 2048 ± 0 8 ± 0 24h 2048 ± 0 80 ± 0 2048 ± 0 8 ± 0 8 1h 2048 ± 0 128 ± 0 32 ± 0 256 ± 0 24h 2048 ± 0 64 ± 0 32 ± 0 256 ± 0 9 1h 2048 ± 0 128 ± 0 2048 ± 0 8 ± 0 24h 2048 ± 0 128 ± 0 2048 ± 0 8 ± 0 10 1h 2048 ± 0 128 ± 0 32 ± 0 256 ± 0 24h 2048 ± 0 128 ± 0 32 ± 0 256 ± 0 11 1h 2048 ± 0 128 ± 0 2048 ± 0 8 ± 0 24h 2048 ± 0 128 ± 0 2048 ± 0 8 ± 0 12 1h 2048 ± 0 64 ± 0 32 ± 0 256 ± 0 24h 2048 ± 0 192 ± 0 32 ± 0 512 ± 0
* MIC value not determinable due to a strange growth pattern on the kanamycin rows of the MIC plate (see Appendix 7) Results for ampicillin are shown in Figure 5. All the transfers yielded a MIC value > 2048 µg/mL, which was the highest used ampicillin concentration in this experiment. MIC results for tetracycline are displayed in Figure 6. The 1h transfers yielded tetracycline MIC values ranging between 64 – 224 µg/mL, with a mean of 133 µg/mL. A similar range and mean were found for the 24h transfers: 64 – 192 µg/mL and 127 µg/mL respectively.
Results for kanamycin are illustrated in Figure 7. All the odd number transfers (transfer 1,3,5,7,9,12) yielded MIC values > 2048 µg/mL, whereas even number transfer (transfer 2,4,6,8,10,12) MIC values ranged between 24 – 32 µg/mL, with a mean of 31 µg/mL. Important to note is the missing value for the 1h incubation samples of transfer 2. In this case we could not determine the MIC value due to a strange growth pattern on the MIC plate (see Appendix 7). Lastly, the MIC values for
chloramphenicol showed a reverse pattern compared to the kanamycin results (Figure 8). For odd number transfers, a range of 7 – 91 µg/mL and a mean of 18 µg/mL were found. The even number transfers yielded a range of 256 – 512 µg/mL with a mean of 373 µg/mL.
Figure 5: The Minimum Inhibitory Concentration (MIC) of ampicillin for the 1h and 24h samples of all transfers. Shown are the mean MIC values, all exceed (>) the highest antibiotic concentration used in this experiment (2048 µg/mL). Figure 6: The Minimum Inhibitory Concentration (MIC) of tetracycline for the 1h and 24h samples of all transfers. Shown are the mean MIC values and standard deviations.
Figure 7: The Minimum Inhibitory Concentration (MIC) of kanamycin for the 1h and 24h samples of all transfers. Shown are the mean MIC values and standard deviations.
* MIC not determinable due to strange growth pattern at high and low concentrations kanamycin, see Appendix 7. Figure 8: The Minimum Inhibitory Concentration (MIC) of chloramphenicol for the 1h and 24h samples of all transfers. Shown are the mean MIC values and standard deviations.
Altogether, the MIC results shown in Figure 5-8 are based on small sample sizes (n = 1-3), which made it difficult to perform a statistical analysis on the results. However, these MIC experiments were primarily intended as control experiments to investigate if the transfers yielded the expected transconjugant (JW3686 or MG1655), and indeed all odd number transfers showed a high MIC for kanamycin and a low MIC for chloramphenicol, whereas this pattern was reversed for even number transfers. Therefore, it can be concluded that for each transfer, the donor successfully transferred its plasmid to the acceptor.
In addition, the resistance for ampicillin and tetracycline was investigated, since it is known that the ESBL 3170.1 plasmid generates resistance against these antibiotics. Results in Figure 5 show that the
ampicillin MIC remained above 2048 µg/mL for all transfers. However, due to the limited MIC range in this experiment, it is impossible to state that the actual MIC value remained stable over all transfers. For tetracycline, all MIC values did fall within the range of concentrations that was used in this experiment (Figure 6). MIC values deviating around 128 µg/mL were found. Since no clear pattern is visible in the data, it can be concluded that the tetracycline resistance roughly remained stable over all transfers.
Sequencing results show nearly no changes in the plasmid after 12 transfers Identification of the plasmid using PlasmidFinder 2.0
In order to investigate whether the three plasmids isolated from the 1h samples of transfer 12 were indeed the same plasmids as the three plasmids from the 1h samples of transfer 1, the FASTA files of the six plasmids were uploaded in PlasmidFinder. After searching for similar plasmids in the database, PlasmidFinder returned the IncI1-I(gamma) plasmid as most comparable plasmid with 99.3%
similarity for all samples (threshold for minimum% identity: 95%; minimum % coverage: 60%). From this can be concluded that the plasmid was successfully transferred from transfer 1 until transfer 12.
Identification of acquired antimicrobial resistance genes using ResFinder 3.2
To determine which antimicrobial resistance genes were present on the six plasmids, the FASTA files were then uploaded in Resfinder. After searching the plasmids for resistance genes, ResFinder consistently returned three resistance genes: blaCTX-M-1 (ß-lactam resistance gene), tet(A) (Tetracycline resistance gene) and sul2 (Sulphonamide resistance gene), all with 100% similarity (threshold for %ID: 90%, minimum length: 60%). From this can be concluded that the plasmid contains three well-known resistance genes that do not undergo rough changes after 12 transfers.
Investigation of changes in the annotated genes of the plasmid
To investigate the sequence of the annotated genes on the plasmids, the annotated FASTA files were opened in CLC Main Workbench, which revealed four annotated antibiotics resistance genes in all the six plasmids, including a Class A beta-lactamase gene, two tetracycline resistance genes and one
sulphonamide resistance gene. For each of the three samples, the annotated genes on the plasmids
of transfer 1 and transfer 12 were then aligned in order to investigate if any mutations had occurred after the plasmid was transferred 12 times.
Interestingly, the sequences of nearly all genes including the previously mentioned four antibiotics resistance genes showed a 100% similarity between the transfer 1 and transfer 12 samples. This implicates that most genes are strictly conserved, even after 12 plasmid transfers. An exception of this was the gene IncI1 plasmid conjugative transfer pilus-tipadhesin protein PilV. In all the samples of transfer 1, we found the PilV gene with an approximate length of 1,300 base pairs (Figure 9A-C), and we additionally found certain A’ and A segments in transfer 1 sample 3 (Figure 9C), described earlier by Turner et al. (2014). Interestingly, all the transfer 12 samples showed an inverse extension of the
PilV gene accompanied by the A’ and A segments in a specific order: first the A’ segment in the same
direction as the gene (pointing downstream), and then the A segment pointing toward the inverse direction compared to the gene (Figure 9D-F). This will be addressed in the discussion section.
B A
C C
Figure 9: The PilV gene at transfer 1 and transfer 12.
Shown is the PilV gene (long green arrow) and the A’ and A segments (red arrows). A) transfer 1 sample 1 B) transfer 1 sample 2 C) transfer 1 sample 3 D) transfer 12 sample 1 E) transfer 12 sample 2 F) transfer 12 sample 3. Pictures have been made in CLC Workbench (Version 8.1; QIAGEN).
DISCUSSION
In this study, we investigated the effect of multiple plasmid transfers on transfer efficiency and antibiotics resistance of E. coli, and on the sequence of the plasmid itself. To summarize, we found that multiple plasmid transfers do not drastically increase or decrease the transfer efficiency of E. coli. In fact, we detected a high variance of CFU/mL values between transfers after both 1h and 24h incubation. We also conducted MIC experiments that revealed a high MIC value for ampicillin at each transfer, always extending above the highest concentration used in our experiment. In contrast, we found a lower but relatively stable level of tetracycline MIC in all transfers. The MIC patterns of kanamycin and chloramphenicol confirmed that we consistently ended up with the right
transconjugant after each transfer. Lastly, sequencing results showed that all annotated genes at transfer 1 remained intact until transfer 12, except for the gene PilV. Based on our results, we can conclude that multiple plasmid transfers have no effect on transfer efficiency and antibiotics resistance of E. coli, and that nearly all known genes remain strictly conserved.
With regards to our transfer efficiency results, there are several explanations for the high variation of CFU/mL values we found. First of all, we used two different E. coli strains in this experiment, which invariably induced variation in transfer efficiency between the odd and even number transfers. Interestingly, this difference was much clearer in the transfers that had incubated for 24 hours, most probably because the JW3686 strain survives particularly better after 24h glucose deprivation
compared to the MG1655 bacterium. Similarly, the fact that most transconjugants generally showed a higher CFU/mL after 1h incubation compared to 24h incubation could be explained by the death of a substantial part of the transconjugants after 24 hours glucose deprivation instead of 1 hour.
E
F D
Besides these biological explanations for the variance in our dataset, some methodological factors might have contributed to the variance in our dataset as well. First of all, we consistently took one colony from the 1h transconjugant plate of each sample to use as donor for the next transfer. Since we did not use any selection criterium for this colony, it is possible that we sometimes picked a particularly weak or strong colony as donor for the next transfer, leading to a higher or lower transfer efficiency of the next transconjugant. However, we found no support for this hypothesis when we corrected for the donor and acceptor (Appendix 5). Another variation inducing factor might the 0-300 count range we used instead of the recommended range of 30-300 (Breed and Dotterer, 1916). However, we wanted to use this range because several transfers yielded less than 30 CFU. Furthermore, an important contributor to the variation might be the capability of the bachelor student that conducted the experiments. After transfer 2, all experiments were executed without direct supervision, which could explain the divergent results of transfer 3-8. Since the results of transfer 9-12 biologically look more plausible, this could be due to the increased skills of the student as the project progressed. However, since the results of Ivan Fung showed high variations as well, it is also likely that suboptimalities in the experimental set-up have contributed to the high variance in our dataset. For instance, we never used real replicate plates for the transconjugant, but always took the mean of the 10-0, 10-1 and 10-2 dilution results, although it is unclear if the dilution factor is indeed inversely proportional to the amount of CFU on agar plates.
For some of the even number transfer MIC experiments, we detected growth at higher
concentrations of kanamycin than the MIC, making it difficult to determine the actual MIC value (Appendix 6, Figure 7 *). According to literature, these findings might be explained by a mechanism called the Eagle effect (Prasetyoputri et al., 2019). A previous study postulated that high
concentrations of the quinolone nalidixic acid inhibits the E. coli protein synthesis so drastically that Reactive Oxygen Species (ROS) formation is reduced, ultimately leading to survival of the bacterium at nalidixic acid concentrations high above the MIC (Luan et al., 2018). Therefore, it would be interesting to investigate in future research if the levels of ROS change in E. coli during subjection to high kanamycin concentrations. Furthermore, it would also be interesting to redo our MIC
experiments of ampicillin with higher concentrations in order to identify the real MIC values of ampicillin, and to determine if these MIC values change as the transfers proceed.
Due to some methodological failures that occurred during the plasmid isolation procedure, we were surprised by the good quality of the sequencing results. For instance, the three samples of transfer 12 were grown overnight without shaking due to problems with the shaking incubator, while the
protocol recommended shaking (Appendix 1, day 2 step 3). Nevertheless, the sequence data was of sufficient quality for further analysis, and we found that nearly all annotated genes on the plasmid were 100% identical between transfer 1 and transfer 12. Therefore, we can state that almost all genes are strictly conserved after 11 plasmid transfers. However, it is important to note that this study only focused on the sequence of genes, implicating that non-coding regions or promoters could actually undergo changes after multiple plasmid transfers. Nonetheless, we found no obvious changes in transfer efficiency and antibiotics resistance as the transfers proceeded, suggesting that dramatic changes in non-coding parts of the plasmid are not very probable.
Interestingly, PilV was the only gene on the plasmid we found to be dissimilar between transfer 1 and transfer 12. According to literature, the protein PilV adhesin binds to lipopolysaccharide on recipient bacteria and hence plays an important role in bacterial mating (Ishiwa & Komano, 2004). Previous research revealed seven possible rearrangements of this gene in the IncI1 plasmid R64, of which only one enabled successful mating in E. coli (Turner et al., 2014). PilV is also known as Shufflon protein A (UniProt, 2019) and only becomes a functional protein when an A’ segment is added on the end of
the gene, followed by an A segment pointing in the inverse direction (Appendix 3). Interestingly, this is exactly the difference we found between the transfer 1 and transfer 12 samples. Although sample 3 of transfer 1 did contain the A and A’ segments, they were placed in an inactive order (first A, then A’ inverse, Figure 9C), whereas all the transfer 12 samples were placed in the active order (first A’, then A inverse). These findings suggest that multiple plasmid transfers could activate PilV by genetic rearrangement, but since we found no improvement in the transfer efficiency as the transfers
proceeded, this might be of minor significance in the conjugation process. In future research, it would be interesting to investigate at which specific transfer (2-11) the PilV rearrangement occurs.
Altogether, we demonstrated for the first time that a plasmid can be transferred 12 times without striking effects on the phenotype of E. coli, nor on the genotype of the plasmid. This underlines the robustness and stability of the conjugation process, and demonstrates the prominence of conjugation in the development of antibiotics resistance (Nazarian et al., 2018). Hopefully, future research will reveal new characteristics and potential interference points in conjugation, making it possible to target this important transmission mechanism, for instance in the human food chain. For now, antibiotics resistance remains an increasing threat just as the WHO predicted one year ago (WHO, 2019). However, unlike the current COVID-19 pandemic, we fortunately still have time to develop new strategies in order to prevent an antibiotics resistance peak in the coming decades.
ACKNOWLEDGMENTS
I would like to thank prof. dr. B.H. ter Kuile for adding me to his research group, and T.S. Darphorn (MSc) for being a patient and always available supervisor during the internship. I also want to thank my colleague M. Nulkes (student) for conducting my last MIC experiments.
LITERATURE LIST
Breed, R.S. & Dotterer, W.D. (1916). The Number of Colonies Allowable on Satisfactory Agar Plates.
Journal of Bacteriology, 1,321-331.
Händel, N., Otte, S., Jonker, M., Brul, S., & Ter Kuile, B.H. (2015). Factors That Affect Transfer of the lncl1 ß-Lactam Resistance Plasmid pESBL-283 between E. coli Strains. PLoS One, 10,e0123039. Harrison, E. & Brockhurst, M.A. (2012). Plasmid-mediated horizontal gene transfer is a coevolutionary process. Trends in Microbiology, 20, 262-267.
Hazan, R., Que, Y., Maura, D., & Rahme, L.G. (2012). A method for high throughput determination of viable bacteria cell counts in 96-well plates. BMC Microbiology, 12,259.
Ishiwa, A. & Komano, T. (2004). PilV adhesins of plasmid R64 thin pili specifically bind to the lipopolysaccharides of recipient cells. Journal of molecular biology, 343,615-625.
Landers, T.F., Cohen, B., Wittum, T.E., & Larson, E.L. (2012). A Review of Antibiotic Use in Food Animals: Perspective, Policy and Potential. Public Health Reports, 127,4-22.
Leverstein-van Hall, M.A., Dierikx, C.M., Cohen, S.J., Voets, G.M., Van den Munckhof, M.P., Van Essen-Zandbergen, A., et al. (2011). Dutch patients, retail chicken meat and poultry share the same ESBL genes, plasmids and strains. Clinical Microbiology and infection, 17,873-880.
Levy, S.B., FitzGerald, G.B., & Macone, A.B. (1976). Spread of antibiotic-resistant plasmids from chicken and from chicken to man. Nature, 260,40-42.
Luan, G., Hong, Y., Drlica, K., & Zhao, X. (2018). Suppression of Reactive Oxygen Species Accumulation Accounts for Paradoxical Bacterial Survival at High Quinolone Concentration. Antimicrobial Agents
and Chemotherapy, 62.
Mazel, D., & Davies, J. (1999). Antibiotic resistance in microbes. Cell and molecular life sciences, 56,742-754.
Medaney, F., Dimitriu, T., Ellis, R.J., & Raymond, B. (2016). Live to cheat another day: bacterial dormancy facilitates the social exploitation of ß-lactamases. The ISME Journal, 10,778-787. Millan, A.S. (2018). Evolution of Plasmid-Mediated Antibiotic Resistance in the Clinical Context.
Trends in Microbiology, 26,978-985.
Nazarian, P., Tran, F., & Boedicker, J.Q. (2018). Modeling Multispecies Gene Flow Dynamics Reveals the Unique Roles of Different Horizontal Gene Transfer Mechanisms. Frontiers in Microbiology, 9,2978.
Norman, A., Hansen, L.H., & Sørensen, S.J. (2009). Conjugative plasmids: vessels of the communal gene pool. Philosophical Transactions of the Royal Society B: Biological sciences, 364,2275-2289. Prasetyoputri, A., Jarrad, A.M., Cooper, M.A., & Blaskovich, M.A.T. (2019). The Eagle Effect and Antibiotic-Induced Persistence: Two Sides of the Same Coin? Trends in Microbiology, 27,339-354. Thomas, C.M. & Nielsen, K.M. (2005). Mechanisms of, and barriers to, horizontal gene transfer between bacteria. Nature reviews. Microbiology, 3,711-721.
Turner, P.E. Williams, E.S.C.P., Okeke, C., Cooper, V.S., Duffy, S., & Wertz, J.E. (2014). Antibiotic resistance correlates with transmission in plasmid evolution. Evolution, 68.
Uniprot. (2019). UniProtKB – A0A0S0ZQ18 (A0A0S0ZQ18_SALEN). On internet: https://www.uniprot.org/uniprot/A0A0S0ZQ18, consulted on 28-03-2020.
Van Boeckel, T.P., Brower, C., Gilbert, M., Grenfell, B.T., Levin, S.A., et al. (2015). Global trends in antimicrobial use in food animals. Proceedings of the National Academy of Sciences of the United
States of America, 112,5649-5654.
World Health Organization. (2019). New report calls for urgent action to avert antimicrobial resistance crisis. On internet: https://www.who.int/news-room/detail/29-04-2019-new-report-calls-for-urgent-action-to-avert-antimicrobial-resistance-crisis, consulted on 19-08-2019.
You, Y., & Silbergeld, E.K. (2014). Learning from agriculture: understanding low-dose antimicrobials as drivers of resistome expansion. Frontiers in Microbiology, 5,284.
Appendix 1: QIAGEN plasmid isolation procedure
Qiagen Plasmid Maxi Kit
Before you start:
-
Add RNase solution to buffer P1 (use 1 vail per bottle, briefly centrifuge, final
concentration 100 μg / ml). Store it at 4 ° C
-
Check buffer 2 for precipitation, if this is so, heat at 37 ° C (incubator).
-
Put buffer P3 at 4 ° C.
-
Heat elution buffer QF to 65 ° C, start heating up as soon as you start the buffer steps.
-
Make sure you stick to the maximum culture volumes, otherwise this leads to an
inefficient cell lysis.
Preparation of LB medium:
10 g of Trypton
5 g Yeast Extract
10 g NaCl
1000 ml H₂O
Prepare Erlenmeyer flasks with 5 ml LB medium and 400 ml LB medium. You can keep this
sterile just refrigerated without antibiotics.
Requirements (per sample!):
Erlenmeyer 20ml, 500 ml and 1 liter (sterile)
1 500 ml sterile bucket (400 ml in it)
Tips 1000 and 200 μl
50 ml conical tubes
Performance:
Day 1
1.
Spread the desired clone on a freshly prepared LB plate with antibiotic marker.
Day 2
2. Innoculate 1 colony in 5 ml LB medium with antibiotic marker (in 20 ml sterile Erlenmeyer
flask) and grow to a late logarithmic phase in about 8 hours at 37 ° C (shaker).
Possibly: measuring OD600 before and after incubation of theculture, then you have better insight in growth cells.
3. Inoculate 0.4 ml of this starter culture in 400 ml LB with antibiotic marker and grow in
about 12-16 hours to a late logarithmic / stationary phase at 37 ° C (shaker). This can be
overnight.
Day 3
4. Fill 400 ml bucket with your mature culture. Do 2 samples at the same time then you have
counterweight in centrifuge. Centrifuge at 6000 rpm for 15 min in the Sorvall centrifuge.
Label your buckets!
The estimated amount of cell pellet will be about 1.5 grams (assuming 500 ml of culture)
5. First pipette 10 ml of supernatant into a 50 ml tube, keep your pipette for a while.
6. Carefully remove the supernant by pouring and keeping the cell pellet at the top of the
bucket. Catch it up again in the Erlenmeyer flask containing your starter culture = waste!
7. Now pipette your 10 ml supernatant back into the bucket, resuspend your pellet here (mix
well) and then transfer it to your 50 ml tube. And another spin in Sorvall centrifuge 6000 rpm
for 10 min.
Lower rpm does not stick pellet firmly to the wall.
8. Remove all supernatant with a pipette or the suction pump and freeze the pellet at -20 ° C
(if you don’t wish to continue immediately).
Day 4
First put your Buffer QF down at 65 ° C to warm up!
9. Thaw the pellet on ice and resuspend your pellet with 10 ml P1 buffer by pipetting or
vortexing.
10. Add 10 ml of buffer P2, gently shake back and forth 4-6 times and incubate at room
temp. For 5 min. (Not longer!)
Close P2 buffer immediately because of CO₂.
11. Add 10 ml of buffer P3, gently shake back and forth 4-6 times and incubate on ice for 20
min.
If the mixture remains viscous then continue mixing.
12. Centrifuge the tubes at 10,000 rpm for 60 min. (Protocol says> 20,000 for 30 min.).
Before mixing the tube, mix again. After centrifugation, the supernatant should be clear,
transfer it to a new 50 ml tube.
13. Centrifuge the tubes again at 10,000 rpm for 30 min. (Protocol says> 20,000 for 15 min.).
If your supernatant is still too cloudy, your filter can get full.
14. Meanwhile, flush the Qiagen tip 500 with 10 ml of QBT buffer, collect it in an Erlenmeyer
flask (= waste)
15. Now place the Qiagen tip 500 on a erlenmeyer waste flask.
16. Load the remaining clear lysate on your Qiagen tip 500 and let it run through.
Is your lysate cloudy again then first centrifuge or filter again.
17. Now wash your Qiagen tip 500 with 2x 30 ml Buffer QC.
18. Put the Qiagen tip 500 back on a clean 50 ml tube.
19. Add 15 ml of the preheated Buffer QF and let it run through.
20. You can now store the eluate overnight (but no longer) at 4 ° C.
Day 5
First put your elution buffer at 55 ° C to warm up!
21. Remove the eluate from the refrigerator and allow to reach room temperature, add 10.5
ml of isopropanol (also to room temp.). Mix this and then immediately centrifuge at 5000
rpm for 60 min at 4 ° C. First mark on the outside of the tube where you expect the pellet
because that is rather difficult to see after centrifugation.
22. Remove the supernatant very carefully because the pellet is loose.
23. Wash the pellet with 5 ml of 70% ethanol (in room temp.) And centrifuge at 5000 rpm for
60 min at 4 ° C.
24. Remove the supernatant and allow the pellet to air dry for 5-10 min. No longer does the
pellet dissolve properly!
25. Dissolve the pellet in 0.5 ml elution buffer (10mM Tris-CL pH 8.5). Do not pipette up and
down, it will damage the DNA.
26. Store the epje with the purified plasmid DNA at -20 ° C
Same day 5 or another day
27. Measure the concentration of DNA with the nanodrop at 260 nM
28. Run a 1% agarose gel. Make a gel from 100 ml 1x TAE, 1 gram agarose and 5 μl
Midorigreen. Put a comb with small clasps (14x) and let it solidify for 30 minutes.
29. Take 5 μl of each sample and add 3 μl of loading dye 6x. Load this on the gel and take a
DNA 1 Kb + ladder (10 μl) for verification.
Appendix 2: Ethanol precipitation protocol
Ethanol precipitation – for concentration and cleaning of samples
Before starting have the following ready:
- 96 % ethanol should be cold (bottle stored at -20 degrees Celsius). - Use TE as elution buffer and put this at 55-60 degrees.
- Sodium acetate and 70 % ethanol can be kept at room temperature.
- Sodium acetate (pH 5.2, 3M) is made as follows: 24.6 g Sodium acetate (N20 or 21) is dissolved in 70 ml milliQ, pH is adjusted to 5.2 using glacial (=100%) acetic acid (A406). Add water up to 100 mL, mix and done.
- Small bucket of ice
1. Defrost samples on ice.
2. For concentration take as much of the sample as you want (200-400 μl) and put it in a clean tube. For cleaning the samples use everything (20-30 μl), you can reuse the same tube. 3. Add 1/10 volume of sodium acetate (pH 5.2, 3M). For example: 200 μl sample = 20 μl SA. 4. Add 3 volumes of ice cold 96 % ethanol and mix gently by pipetting slowly up and down. For
example: 200 μl sample = 600 μl ethanol. 5. Incubate for 15 minutes on ice (minimum).
6. Mark where your pellet will be and centrifuge for 30 minutes at room temperature on the highest rpm/rcf.
7. Discard the supernatant very carefully, keeping an eye on your pellet. 8. Wash the pellet gently with 70 % ethanol (200 a 300 μl).
9. Spin again in the centrifuge for 15 minutes at room temperature on the highest rpm/rcf. 10. Discard the supernatant very carefully, keeping an eye on your pellet.
12. Add 20-30 μl warm TE elution buffer, pipette carefully up and down to elute and mix the DNA.
13. Measure amount of DNA with the nanodrop. Ideal is 260/230 and 260/280 of 1.8. Try not to go underneath 1 and have more than 20 ng/ul.
Appendix 3: Primer sequences and sequential order of
segments within the PilV gene (Turner et al., 2014)
Figure from Turner et al. (2014).The active form of the PilV gene, called pilVA’ is characterized by an A’ segment in the same direction as the gene, and an A segment in the inverse direction.
Primer of A segment: 5’ GGTTCACATAGAGGTTCATTCTCAGGGC 3’ Primer of A’ segment: 5’ GCAATTCGACTTCTGTGCCATTGCC 3’
Primers of the B, B’, C, C’ and D’ segments are also described in this paper. Some of these primers also aligned to the plasmid we used in our study, but not with a consistent difference between transfer 1 and transfer 12 as we saw for the A and A’ segments. An explanation for this could be that the primers of the A and A’ segments were a 100% match in our plasmid by coincidence, whereas the other segments are slightly different in our IncI1 plasmid and hence did not align well.
Appendix 4: Raw untransformed CFU data for acceptor,
donor and transconjugant
Transfer Transconjugant type Incubation time Acceptor sample 1 mean (CFU/mL) + SD Acceptor sample 2 mean (CFU/mL) + SD Acceptor sample 3 mean (CFU/mL) + SD Log10 mean of samples (CFU/mL) + SD 1 JW3686 1h 1.38E+08 ± 6.00E+02 7.55E+07 ± 2.02E+07 9.88E+07 ±
3.76E+07 1.04E+08 ±2.58E+07
24h 1.82E+07 ±
3.35E+06
2.47E+07 ± 3.06E+06
7.40E+07 ±
1.80E+03 3.90E+07 ±2.49E+07
2 MG165 1h 4.16E+08 ±
0.00E+00
1.62E+08 ± 1.52E+04
9.40E+07 ±
3.60E+03 2.24E+08 ±1.39E+08
24h 4.30E+06 ±
9.81E+05
9.10E+06 ± 2.83E+06
1.13E+07 ±
7.51E+05 8.23E+06 ±2.92E+06
3 JW3686 1h 1.86E+08 ±
0.00E+00
1.86E+08 ± 2.00E+03
1.84E+08 ±
5.79E+02 1.85E+08 ±9.43E+05
24h 8.80E+07 ±
8.99E+02
5.83E+07 ± 2.14E+06
4.00E+07 ±
1.66E+02 6.21E+07 ±1.98E+07
4 MG1655 1h 3.44E+06 ±
4.64E+05
1.08E+07 ± 7.64E+06
3.87E+06 ±
1.49E+06 6.03E+06 ±3.37E+06
24h 5.90E+06 ±
2.25E+06
3.20E+06 ± 6.93E+05
9.52E+06 ±
3.79E+06 6.21E+06 ±2.59E+06
5 JW3686 1h 1.02E+08 ± 8.11E+02 1.48E+08 ±
8.40E+03
1.44E+08 ±
1.00E+02 1.31E+08 ±2.08E+07 24h 9.60E+07 ± 5.77E+02 1.10E+08 ±
5.00E+03
1.66E+08 ±
1.23E+03 1.24E+08 ±3.02E+07
6 MG1655 1h 1.79E+07 ±
2.37E+06
1.55E+07 ± 5.97E+06
6.00E+05 ±
4.24E+05 1.13E+07 ±7.66E+06
24h 0.00E+00 ** 1.38E+06 ±
3.47E+05
5.88E+04 ±
6.00E+04 4.80E+05 ±6.37E+05
7 JW3686 1h 2.30E+08 ±
7.90E+03
2.22E+08 ± 1.00E+02
7.20E+07 ±
2.00E+02 1.75E+08 ±7.27E+07
24h 1.50E+08 ±
1.10E+04
3.83E+07 ± 1.29E+07
5.05E+07 ±
2.47E+06 7.96E+07 ±5.00E+07
8 MG1655 1h 3.56E+06 ±
9.91E+05
3.35E+06 ± 8.95E+05
2.99E+05 ±
1.53E+05 2.40E+06 ±1.49E+06
24h 9.13E+05 ±
5.52E+05
6.33E+05 ± 6.33E+05
1.71E+06 ±
1.15E+06 1.08E+06 ±4.55E+05
9 JW3686 1h 1.50E+08 ±
3.10E+03
1.30E+08 ± 2.20E+03
8.60E+07 ±
24h 6.52E+07 ± 1.89E+07
5.31E+07 ± 1.90E+07
7.76E+07 ±
1.29E+07 6.53E+07 ±1.00E+07
10 MG1655 1h 6.76E+06 ±
4.64E+06
1.24E+07 ± 8.79E+06
5.41E+07 ±
3.10E+07 2.44E+07 ±2.11E+07
24h 2.42E+06 ±
1.62E+06
4.00E+06 ± 1.01E+06
1.89E+06 ±
2.21E+05 2.77E+06 ±8.97E+05
11 JW3686 1h 4.40E+07 ±
1.00E+02
5.38E+07 ± 1.15E+05
5.26E+07 ±
4.24E+05 5.01E+07 ±4.36E+06
24h 2.20E+07 ±
1.32E+02
2.28E+07 ± 9.93E+06
2.94E+07 ±
2.66E+06 2.47E+07 ±3.32E+06
12 MG1655 1h 5.17E+06 ±
9.55E+05
3.51E+06 ± 1.00E+06
8.61E+06 ±
1.72E+06 5.76E+06 ±2.12E+06
24h 3.20E+06 ±
1.20E+06
4.21E+06 ± 8.61E+05
5.02E+06 ±
2.60E+06 4.14E+06 ±7.44E+05
Transfer Transconjugant type Incubation time Donor sample 1 mean (CFU/mL) + SD Donor sample 2 mean (CFU/mL) + SD Donor sample 3 mean (CFU/mL) + SD Mean of samples (CFU/mL) + SD 1 JW3686 1h 3.70E+08 ± 6.00E+02 5.00E+08 ± 3.10E+03 1.52E+08 ± 3.00E+03 3.41E+08 ± 1.44E+08 24h 5.52E+07 ± 6.56E+02 1.76E+08 ± 2.22E+03 2.16E+08 ± 1.80E+03 1.49E+08 ± 6.84E+07 2 MG165 1h 1.40E+08 ± 0.00E+00 1.60E+08 ± 1.52E+04 1.16E+08 ± 3.60E+03 1.39E+08 ± 1.80E+07 24h 5.59E+07 ± 2.25E+06 5.57E+07 ± 5.95E+06 7.42E+07 ± 1.27E+06 6.19E+07 ± 8.67E+06 3 JW3686 1h 4.00E+05 ± 3.53E+05 3.72E+05 ± 2.79E+05 5.73E+05 ± 4.77E+05 4.48E+05 ± 8.88E+04 24h 1.20E+08 ± 8.99E+02 6.90E+07 ± 1.33E+07 1.78E+08 ± 1.66E+02 1.22E+08 ± 4.45E+07 4 MG1655 1h 9.40E+07 ± 1.11E+03 1.22E+08 ± 4.00E+02 6.65E+07 ± 2.17E+07 9.42E+07 ± 2.27E+07 24h 8.40E+07 ± 3.23E+07 3.28E+08 ± 4.71E+01 2.18E+08 ± 4.00E+02 2.10E+08 ± 9.98E+07 5 JW3686 1h 1.96E+08 ± 8.11E+02 1.96E+08 ± 8.40E+03 4.70E+08 ± 1.00E+02 2.87E+08 ± 1.29E+08 24h 8.20E+07 ± 5.77E+02 1.32E+08 ± 5.00E+03 1.12E+08 ± 1.23E+03 1.09E+08 ± 2.05E+07 6 MG1655 1h 5.86E+07 ± 6.58E+06 1.62E+08 ± 1.32E+03 9.00E+07 ± 5.00E+02 1.04E+08 ± 4.33E+07 24h 5.80E+07 ± 7.54E+01 3.60E+07 ± 9.43E+00 1.02E+08 ± 1.30E+03 6.53E+07± 2.74E+07 7 JW3686 1h 1.62E+08 ± 7.90E+03 3.06E+08 ± 1.00E+02 1.48E+08 ± 2.00E+02 2.05E+08 ± 7.14E+07 24h 8.40E+07 ± 1.10E+04 6.04E+07 ± 7.42E+03 4.96E+07 ± 8.27E+03 6.46E+07 ± 1.44E+07
8 MG1655 1h 7.22E+07 ± 1.49E+07 3.40E+07 ±
6.07E+02 5.22E+07 ± 5.66E+06 5.28E+07 ± 1.56E+07 24h 9.00E+07 ± 8.96E+02 5.98E+07 ± 5.89E+06 5.40E+07 ± 3.51E+02 6.79E+07 ± 1.58E+07
9 JW3686 1h 1.20E+08 ± 3.10E+03 1.32E+08 ±
2.20E+03
6.40E+07 ± 1.30E+03
1.05E+08 ±
2.96E+07
1.18E+03 1.84E+03 7.12E+07
10 MG1655 1h 2.20E+08 ± 1.58E+03 1.46E+08 ±
4.23E+03 1.88E+08 ± 2.10E+03 1.85E+08 ± 3.03E+07 24h 1.98E+08 ± 2.00E+03 1.54E+08 ± 2.04E+03 1.40E+08 ± 1.97E+02 1.64E+08 ± 2.47E+07 11 JW3686 1h 1.04E+08 ± 1.00E+02 7.80E+07 ± 3.58E+02 1.70E+08 ± 1.13E+04 1.17E+08 ± 3.87E+07
24h 1.62E+08 ± 1.32E+02 1.07E+08 ± 3.53E+07
1.56E+08 ± 8.84E+02
1.42E+08 ±
2.47E+07
12 MG1655 1h 5.22E+07 ± 4.10E+06 4.72E+07 ±
7.92E+06
4.14E+07 ± 9.48E+06
4.69E+07 ±
4.41E+06
24h 2.42E+07 ± 1.27E+06 4.61E+07 ± 4.56E+06
5.41E+07 ± 6.87E+06
4.15E+07 ±
1.26E+07 * no visible Colony Forming Units on the plates of this sample.
Transfer Transconjugant type Incubation time Acceptor sample 1 mean (CFU/mL) + SD Acceptor sample 2 mean (CFU/mL) + SD Acceptor sample 3 mean (CFU/mL) + SD Log10 mean of samples (CFU/mL) + SD 1 JW3686 1h 1.38E+08 ± 6.00E+02 7.55E+07 ± 2.02E+07 9.88E+07 ±
3.76E+07 1.04E+08 ±2.58E+07
24h 1.82E+07 ±
3.35E+06
2.47E+07 ± 3.06E+06
7.40E+07 ±
1.80E+03 3.90E+07 ±2.49E+07
2 MG165 1h 4.16E+08 ±
0.00E+00
1.62E+08 ± 1.52E+04
9.40E+07 ±
3.60E+03 2.24E+08 ±1.39E+08
24h 4.30E+06 ±
9.81E+05
9.10E+06 ± 2.83E+06
1.13E+07 ±
7.51E+05 8.23E+06 ±2.92E+06
3 JW3686 1h 1.86E+08 ±
0.00E+00
1.86E+08 ± 2.00E+03
1.84E+08 ±
5.79E+02 1.85E+08 ±9.43E+05
24h 8.80E+07 ±
8.99E+02
5.83E+07 ± 2.14E+06
4.00E+07 ±
1.66E+02 6.21E+07 ±1.98E+07
4 MG1655 1h 3.44E+06 ±
4.64E+05
1.08E+07 ± 7.64E+06
3.87E+06 ±
1.49E+06 6.03E+06 ±3.37E+06
24h 5.90E+06 ±
2.25E+06
3.20E+06 ± 6.93E+05
9.52E+06 ±
3.79E+06 6.21E+06 ±2.59E+06
5 JW3686 1h 1.02E+08 ± 8.11E+02 1.48E+08 ±
8.40E+03
1.44E+08 ±
1.00E+02 1.31E+08 ±2.08E+07 24h 9.60E+07 ± 5.77E+02 1.10E+08 ±
5.00E+03
1.66E+08 ±
1.23E+03 1.24E+08 ±3.02E+07
6 MG1655 1h 1.79E+07 ±
2.37E+06
1.55E+07 ± 5.97E+06
6.00E+05 ±
4.24E+05 1.13E+07 ±7.66E+06
24h 0.00E+00 ** 1.38E+06 ±
3.47E+05
5.88E+04 ±
6.00E+04 4.80E+05 ±6.37E+05
7 JW3686 1h 2.30E+08 ±
7.90E+03
2.22E+08 ± 1.00E+02
7.20E+07 ±
2.00E+02 1.75E+08 ±7.27E+07
24h 1.50E+08 ±
1.10E+04
3.83E+07 ± 1.29E+07
5.05E+07 ±
2.47E+06 7.96E+07 ±5.00E+07
8 MG1655 1h 3.56E+06 ±
9.91E+05
3.35E+06 ± 8.95E+05
2.99E+05 ±
1.53E+05 2.40E+06 ±1.49E+06
24h 9.13E+05 ±
5.52E+05
6.33E+05 ± 6.33E+05
1.71E+06 ±
1.15E+06 1.08E+06 ±4.55E+05
9 JW3686 1h 1.50E+08 ±
3.10E+03
1.30E+08 ± 2.20E+03
8.60E+07 ±
1.30E+03 1.22E+08 ±2.67E+07
24h 6.52E+07 ±
1.89E+07
5.31E+07 ± 1.90E+07
7.76E+07 ±
1.29E+07 6.53E+07 ±1.00E+07
Appendix 5: Dependence of transfer efficiency on donor and
acceptor
Dependence of the transfer efficiency on the number of viable acceptor bacteria (CFU).
Shown is the ratio of the mean log CFU of the transconjugants divided by the mean log CFU of the acceptor for each transfer.
Dependence of the transfer efficiency on the number of viable donor bacteria (CFU).
Shown is the ratio of the mean log CFU of the transconjugants divided by the mean log CFU of the acceptor for each transfer.
Dependence of the transfer efficiency on the mean number of viable donor and acceptor bacteria.
Shown is the ratio of the mean log CFU of the transconjugants, divided by the mean of two means: the mean log CFU of the acceptor and the mean log CFU of the donor.
Appendix 6: Colony Forming Units results of Ivan Fung
Transconjugants
Time
(hours) Acceptor (CFU / mL) Donor (CFU / mL)
Transconjugant (CFU / mL)
I II I II I II
T. knockout 3170.1 1st transfer
1 5.30E+06 4.60E+06 6.20E+07 5.80E+07 1.00E+03 0
24 1.00E+08 2.80E+08 5.76E+08 7.80E+08 2.24E+08 2.60E+07
T. MG1655 YFP 3170.1 2nd transfer
1 1.10E+07 1.13E+07 5.44E+06 4.56E+06 2.00E+02 0
24 6.78E+07 3.38E+07 9.40E+08 1.10E+09 2.40E+05 2.40E+05
T. knockout 3170.1 3rd transfer
1 3.36E+06 3.86E+06 3.04E+06 2.72E+06 1.60E+03 1.20E+03
T. MG1655 YFP 3170.1 4th transfer
1 1.00E+07 1.80E+07 8.00E+06 1.20E+07 0 0
24 2.22E+08 2.44E+08 7.36E+08 8.44E+08 6.00E+06 1.80E+07
T. knockout 3170.1 5th transfer
1 2.12E+06 2.84E+06 3.34E+06 3.22E+06 6.00E+02 0
24 9.60E+07 4.60E+07 3.28E+08 2.36E+08 2.38E+08 2.96E+08
T. MG1655 YFP 3170.1 6th transfer
1 5.32E+06 3.48E+06 1.58E+06 2.32E+06 2.00E+02 0
24 4.17E+07 7.93E+07 7.60E+08 5.52E+08 3.40E+05 6.80E+05
T. knockout 3170.1 7th transfer
1 2.88E+06 2.80E+06 4.48E+06 3.26E+06 4.00E+03 1.40E+03
24 N/A N/A N/A N/A 4.82E+08 2.86E+08
T. MG1655 YFP 3170.1 8th transfer
1 3.46E+06 1.42E+06 4.06E+06 2.04E+06 4.00E+02 0
24 1.77E+07 5.95E+07 1.34E+09 1.17E+09 2.80E+05 4.80E+05
T. knockout 3170.1 9th transfer
1 4.62E+06 5.24E+06 3.72E+06 3.42E+06 2.00E+03 2.80E+03
24 2.66E+08 1.14E+08 1.76E+08 9.80E+07 1.96E+08 1.74E+08
T. MG1655 YFP 3170.1 10th transfer
1 3.30E+06 2.48E+06 2.72E+06 3.32E+06 4.00E+02 2.00E+02
24 3.77E+07 2.95E+07 4.50E+08 4.88E+08 3.40E+05 5.40E+05
T. knockout 3170.1 11th transfer
1 3.29E+06 4.45E+06 2.15E+06 1.01E+06 8.20E+03 5.20E+03
24 6.00E+08 4.90E+08 1.34E+08 2.70E+08 4.88E+08 6.22E+08
Table 7 CFU of the transfer events between species. The amount of CFU after transfers between the
two bacteria strains results in transconjugants for 11 transfers. The CFU of the acceptor, donor and transconjugants are shown.
*N/A: The plates were overgrown and the CFU could not be determined.
Figure 9 Log CFU of 11 transfers of transconjugants at 1 hour and 24 hours. In the 1 hour and 24 hours graphs
show the log CFU against the transfers of both strains.
Appendix 7: Strange growth pattern on kanamycin MIC
2048 µg /mL 1024 µg /mL 512 µg /mL 256 µg /mL 128 µg /mL 64 µg /mL 32 µg /mL 16 µg /mL 8 µg /mL 4 µg /mL 0 µg /mL only medium
MIC of transfer 2 sample 1 after 1 hour of incubation.
