Results

Production of viable phage particles

A BioBrick RFC[10]-compatible phagemids

The in vitro production of engineered M13 phages can be achieved with the concurrent use of two constructs: a phagemid, representing the phage genome, and a helper plasmid, encoding the M13 phage capsid proteins. While the phagemid carries the intergenic sequence for encapsulation, the helper plasmid lacks this sequence, instead carrying an ORI and a selection marker. To obtain an “iGEMized” standard phagemid we started from the commercially available construct by ThermoFisher Scietific pTZ19R and we applied the necessary modifications to make it standard-compatible. To briefly sum them up we deleted the MCS insertend on frame in the LacZ sequence and substituted it with the standard prefix and suffix. In this way we obtained a new BioBrick-RFC[10] standard-compatible construct bearing the encapsidation sequence for M13 phage particles, registered as part BBa_K4727002. We then cloned an RFP expression cassette by taking advantage of the unique cut sites and the correct assemble product was verified by gel screening (Figure 1).

Figure 1. Gel screening of the standard phagemid. In this figure, it can be seen the open construct on the left and the digested construct displaying the RFP cassette at 1kb and the backbone at around 2kb

A BioBrick RFC[10]-compatible helper plasmid

To produce viable phage particles, as mentioned above, we not only needed a phagemid, representing the phage genome, but also a construct that encoded the viral proteins, allowing for the synthesis of the capsid. For this reason a helper construct was needed. Of particular interest to us was the M13cp helper plasmid, built by Chasteen et al.. However, this construct was not compatible with the BioBrick standard, as it carries two PstI restriction sites flanking the ORI; furthermore, it lacks the BioBrick prefix and suffix. Given our intention to develop a new, standard-compatible helper plasmid, a new construct was assembled as detailed in the experiments section. Briefly, to eliminate the prohibited restriction sites and, at the same time, to implement the sequences of the prefix and the suffix, the region embedding the ORI and the selection marker was entirely substituted with the standard-compatible backbone pSB3K3.

Figure 2. PCR mutagenesis of pSB3K3 backbone. (a) The original plasmid map of pSB3K3, with a magnification (b) of the downstream region of the suffix where the new MCS has been inserted by mutagenesis.

The first PCR mutagenesis on pSB3K3, required to insert a new MCS bearing the Alw21I and MluI restriction sites (Figures 2a and 2b), successfully performed only on one out of the six amplified samples (Figure 3).

Figure 3. Electrophoresis gel performed after PCR mutagenesis. Starting from the left, 1 kb DNA ladder was loaded, followed by the six experimental samples, of which only Sample 4 showed proper amplification.

Consequently, this unique sample was carried forward in the following procedures. The next step was to take advantage of the new MCS to clone the M13 genome isolated from the M13cp helper plasmid. The resulting plasmid was registered as part BBa_K4727003 (Figure 4a). Its correct construction was confirmed through gel screening (Figure 4b) and sequencing.

Figure 4. Construction of the helper phage. (a) Plasmid map of the final assembled construct, pSB3KM13. (b) Gel screening of pSB3KM13 double digestion with EcoRI and Alw21I restriction enzymes, from which the expected characteristic bands (722, 2591, and 5489 bp) were obtained.

Phage production and transduction

The phage production and purification process still remains a long and time consuming task that needs appropriate instruments and abilities to be carried out. In our research journey we wanted to test the previously described constructs to assess our ability to produce and purify phage particles.

We started by testing separately the standard helper phage and the standardized phagemid with commercially available constructs. Using the protocol and the passages described in the experiments section we started by using our newly introduced phagemid bearing an RFP expression cassette (BBa_I13521), with the helper plasmid by Chasteen et al.. This first attempt was made to assess the ability of the phagemid to be encapsidated inside the phage particle. After producing and purifying the phages we tried to infect E. coli TOP10 F’ (Figure 5). Following the infection the bacteria have been plated on selective LB agar plates and incubated overnight at 37 C. The following day colonies could be seen on the plate with a weak expression of RFP.

Figure 5. Infection in E. coli TOP10 F’.

This result is not conclusive in a comprehensive manner, for matters of time no further assessment could be done but it will be our priority in the future.

The second attempt was carried out using both of our newly assembled constructs. Again we started with the transformation of the phagemid carrying the RFP expression cassette into competent cells bearing the standard helper phage. After two overnight incubations, a red colony could be seen. This low transformation rate can be explained by the high metabolic burden carried by both the helper plasmid and the phagemid. This colony has been picked and pre-inoculate to produce phage particles as detailed in the experiments section. After purifying the phages no time was left for the transduction assay.

Specificity assay

As we decided to work on a delivery system for our interference mechanism, and given the aim to use it as a therapeutic approach it was our great concern to assess the specificity of the delivery system, namely of the bacteriophage particles. Whereas literature research was clear about the precise tropism of phage particles we wanted to assess ourselves with our engineered viruses to verify if significant discrepancies could be noticed.

For these reasons we set up a simple, yet relevant, specificity assay. Using the transduction protocol (see experiments section), we infected both E. coli F’ as a positive control and E. coli TOP10. It would have been ideal, to further assess the specificity of the three other non-model organisms we worked with, to have a broad understanding of the phenomenon but a simple, yet relevant, issue needed to be considered. All of the three ATCC bacterial strains that we chose (A. baumannii 19606, K. pneumoniae 13883 and P. aeruginosa 10145) were intrinsically resistant to ampicillin. As our phagemid carries a AmpR gene it does not represent an ideal experimental setup to infect these bacteria with the phage particles and subsequently plate them in a LB with Ampicillin plate as they will grow either way.

While time constraints did not allow further investigations, we aim in the future to reiterate the engineering cycle by changing the phagemid resistance (to Chloramphenicol) and perform the above-mentioned assay.

The infected E. coli TOP10 didn’t grow on LB selective plates meaning that the transduction couldn’t occur. Whereas, as expected A. baumannii extensively grew in the plate due to its intrinsic resistance to amplicillin, but no RED colonies could be seen.

Figure 6. Left: infection of A. baumannii. Right: infection of E. coli TOP10

Helper plasmid expressing new tropism genes

The expression cassettes were designed as described in the delivery design page and in vitro synthetised by IDT Technologies. After that they were cloned into the helper plasmid pSB3KM13, previously obtained as described in section above, and the resulting construct was then transformed into the E. coli MG1655 Z1 cell line, which is characterized by the constitutive expression at high levels of the regulatory genes lacI, araC, and tetR. Consequently, this allowed the utilization of the regulated cassettes under suitable promoters in presence of the correct inducer, in this case IPTG. The correct insertion of the tropism expression cassette was assessed via gel electrophoresis, but unfortunately none of the samples resulted positive in our experiments. However, once the helper plasmid expressing a new tropism gene will be obtained, in order to produce phage particles with altered tropism it will be sufficient to co-transform it with the new standard phagemig. The phagemid, due to the presence of the encapsidation sequence, can be inserted into the assembled phage particles. Consequently, it will be possible to generate phages carrying a desired genome with modified tropism.

In conclusion, our modifications propose a versatile platform for the simple and consistent production of phage particles, with the potential to easily modify tropism through interchange of standardized genetic components. Tropism-determining genes will become standardized parts, readily exchangeable among them, thanks to the new helper plasmid pSB3KM13. This platform, if combined with a phagemid of desire, can allow the production of engineered bacteriophages carrying a genome of desire, able to infect specific bacteria of interest.

Future perspectives

In order to apply this work to clinical practice, the consequent logical step will be the purification of bacteriophages with altered tropism and the assessment of their ability to target the desired cells. A possible experimental setup could see the production of phage particles bearing a phagemid carrying an adequate reporter gene (e.g. RFP) under a strong promoter suitable for the target bacterial species. This would allow to easily demonstrate the phage infection, as the bacteria will show the phenotype given by the reporter gene. Once the ability to transduce bacteria will be demonstrated, the reporter gene could be replaced with the desired system to interfere with the genes involved in AMR.

It is remarkable to notice that the helper plasmid pSB3KM13 we developed can represent a versatile platform to easily alter the tropism and to rapidly test new proposed variants. This strategy could specifically gain importance when considering non-characterized pathogens and new strains. It is important to point out that, even if the two proposed genes will work as intended, such an approach could be further explored by combining methods of rational design and random mutation complemented with a massive screening of the generated proteins. Indeed, thanks to the possibilities offered by the advancements in DNA synthesis technologies and their increasingly affordable prices, it will be possible to rapidly generate gene libraries for tropism determinant genes. These libraries can be obtained through random mutations in controlled regions, which can subsequently be easily screened using this platform. By utilizing the helper plasmid, the genes for tropism can be readily inserted and cloned into packaging cells to create phage libraries that express a diverse range of tropism determinants. These libraries can then be swiftly screened against the target pathogen to identify the clone expressing the right protein which, once isolated, will be used to produce phage particles carrying the desired genome that showed efficacy in treating the specific bacterial infection. In the context of an ongoing battle against evolution in order to effectively fight antibiotic resistance, this platform could serve as a promising starting point. We think that the standardization of parts combined with the capability of new generation gene-synthesis can lead to an easy constant update of the platform here developed, making research in this field easier and less time-consuming than before.

dCas9 standardized expression cassette

PCR mutagenesis results

To obtain a cassette compatible with the iGEM standards, starting on the construct of part BBa_J107202, a PCR mutagenesis was carried out to eliminate the EcoRI site while maintaining the codon usage in E. coli. The last thymine of the ATT codon, encoding an isoleucine, has been replaced by a cytosine. The protocol performed was the Quick Protocol for Q5® Site-Directed Mutagenesis Kit. The starting concentration of the miniprep Template DNA was 43.6 ng/μl; so, the plasmid DNA was diluted 1:10 in nuclease free water and then added to the reaction mix. Additionally, during the annealing step of the 25 PCR cycles, a temperature gradient has been tested since the 65°C temperature (suggested by NEB Tm Calculator™) did not lead to any amplification of the plasmid J116-dCas9_3k3 [1] (as verified by gel electrophoresis analysis). As seen in figure 1.A, only the temperatures of 60°C and 58°C led to amplification, and it was decided to use the sample from the former in the subsequent KLD reaction because it had a better quality. This step was successful: after transformation, E. coli TOP10 colonies grew in both plates with antibiotic kanamycin (Figure 1.B); the sequencing of the products obtained from the MiniPrep of some colonies revealed one plasmid, named J116-dCas9_m_3k3, positive for the mutagenesis (Figure 1.C).

FIGURE 1| PCR mutagenesis results. A) Gel electrophoresis screening, from left to right: the 1kb GeneRuler, the blank, the PCR sample at 63°C (no amplification), at 60°C (amplification with few non specific bands), at 58°C (amplification with nonspecific bands below the expected one at about 7 kb); B) The colonies grown on the LB+kanamycin plates [25 μg/mL], (left) 100 μl of E.coli TOP10 transformed with the product of the KLD reaction were plated; (right) the result of the concentration of the bacterial suction from the first one. C) Alignment between the original sequence of dCas9 around nucleotide 1339 5'-end and the sequencing of the colony 1C (C replaces T).

Despite the optimization of the temperature and of some steps of the protocol, the setup of the experiment has been simple and ultimately successful: a standardized expression cassette was obtained for dCas9 (dCas9_m), which can be inserted into plasmids exploiting the BioBrick RFC Standard [10]. The new coding sequence for dCas9 has been registered as part BBa_K4727008.

Assessing silencing efficiency

Once introduced the silence mutation the coding sequence of the dCas9 we wanted to assess if any significant alteration was measurable in the silencing efficacy of the dCas9 coded by the mutated sequence. For this a plate reader experiment was set up with the following specifications. Escherichia coli TOP10 chemically competent cells were co-transformed with three plasmids bearing: (a) BBa_I13521 expression cassette, (b) HSL inducible pTET guide RNA (from Bellato et al. Frontiers in Bioeng e biotech, 2022), (c) dCas expression cassette. Two different transformations were made, with the only difference of the dCas construct: in one case, the newly mutated and in the other the one of part BBa_J107202. Different HSL inductions of the pTET-targeting guide have been provided to demonstrate silencing efficacy. Average RFP/OD and Synthesis rate per cell (Scell, as described in [Kelly JR et al., Journal Biol Eng, 2009. doi: 10.1186/1754-1611-3-6]) over the exponential growth phase were computed to draw the following curves.

From this analysis it is possible to infer that no significant difference can be seen in the silencing efficacy of the mutated sequence of the dCas9. Overall both the samples do not reach complete silencing of the RFP expression levels probably due to the high copy number of the plasmid bearing the BBa_I13521 expression cassette. At the same time, even at no levels of HSL induction (0) the expression of RFP is lower than the reference level, this can be explained by the leakage activity of the pLux inducible promoter.

sgRNA sequences and expression cassette

As described earlier in the sgRNA design section of the Wiki, it was decided to synthetize the sgRNA sequences as standardized expression cassette flanked by the Prefix and the Suffix for future clonings, as well as by 6 random bases or more on each side for enzyme landing (Figure 2). The scaffold is the same for all the strain of interest (developed by Qi et al., 2013 [2]). To gain greater efficiency, an additional Terminator rrnBT1 was copied downstream the scaffold, and finally, the J23119 promoter, that we found it was used in all the strain of interest, was inserted exactly upstream of the 20 nt sequence.

FIGURE 2| Standard expression cassette synthesized for every sgRNA used in the project; the 20 nt of the guide are specific for the target gene.

It was decided to use 20 nt long spacers as it’s the length with the best trade-off between efficiency and low off-target score (Cui et al., 2018 [3]); their design was created by running the target gene sequences in the CHOPCHOP, Cas-Designer, CRISPOR and Benchling tools. The analysis of the results and the relative most important parameters was made on the basis of the considerations set out in the design and experiments section of this Wiki, searching for two appropriate guides for each gene. The steps carried out were the same for all the genes of the different strains.

We designed two guides for each target gene, that are ompA (sg157, sg211), orf3 (sg30, sg110), ampC (sg56, sg148), algC (forward - sg36, sg173 - and reverse - sg36, sg98 - CDSs) for A. baumannii, ompA (sg125, sg222) and blaSHV-1 (sg85, sg181) for K. pneumoniae, pvdA (sg210) and ampC (sg119) for P. aeruginosa. The guides can be used individually in the silencing test or together to increase efficiency. The chosen spacer sequences (named as “sg” followed by a number stating for their binding position after the TSS) are shown in Table 1 in correlation with all the parameters that were considered (experiments section of this Wiki).

Table 1| Spacers with their relative parameters.

All off-target sequences were aligned with the genomes of the bacteria of interest and none revealed an actual off-target, except sg56; the putative off-target aligned in an unannotated CDS region with 4 mis-matched (which weren’t in the seed region) and the PAM was non-canonical (CAG). The CDS was analyzed on BLAST selecting “A. baumannii ATCC 19606”, with the resulting hits being two hypothetical proteins, one of which had the length equal to the number of AA (120) of the translet putative off-target (NCBI Reference Sequence: WP_000881947.1). After these considerations, the choice of the sg56 guide is in any case confirmed.

The guides library has been registered in the registry of standard biological parts as follows:

Guide Part name
sg30 orf3 BBa_K4727101
sg36 algC_FW BBa_K4727102
sg36 algC_RV BBa_K4727103
sg56 ampC BBa_K4727104
sg85 bla-SHV1 BBa_K4727105
sg110 orf3 BBa_K4727106
sg125 ompA KP BBa_K4727107
sg148 ampC BBa_K4727108
sg157 ompA BBa_K4727109
sg173 algC FW BBa_K4727110
sg181 bla-SHV1 BBa_K472711
sg211 ompA AB BBa_K4727112
sg222 ompA KP BBa_K4727113
sg98 algC RV BBa_K4727114

Reference

  1. Bellato, M., Frusteri Chiacchiera, A., Salibi, E., Casanova, M., De Marchi, D., Castagliuolo, I., Cusella De Angelis, M. G., Magni, P., & Pasotti, L. (2022). CRISPR Interference Modules as Low-Burden Logic Inverters in Synthetic Circuits. Frontiers in Bioengineering and Biotechnology, 9, 743950. https://doi.org/10.3389/fbioe.2021.743950.
  2. Qi, L. S., Larson, M. H., Gilbert, L. A., Doudna, J. A., Weissman, J. S., Arkin, A. P., & Lim, W. A. (2013). Repurposing CRISPR as an RNA-Guided Platform for SequenceSpecific Control of Gene Expression. Cell, 152(5), 1173–1183. https://doi.org/10.1016/j.cell.2013.02.022.
  3. Cui, Y., Xu, J., Cheng, M., Liao, X., & Peng, S. (2018). Review of CRISPR/Cas9 sgRNA Design Tools. Interdisciplinary Sciences: Computational Life Sciences, 10(2), 455–465. https://doi.org/10.1007/s12539-018-0298-z

BBa_K4727000

To engineer the pSGAb-km plasmid into a standardized backbone according to the BioBrick RFC[10], a digestion reaction was performed to remove the J23119 promoter (containing a SpeI restriction site), the original sgRNA cassette and the Multiple Cloning Site. As explained in the experiments section of this Wiki, the restriction reaction was set up using the HindIII (Thermofisher) and SacI (Thermofisher) restriction enzymes. Then, a properly designed insert (Figure 1), synthesized by Eurofins Genomics, was cloned into the linearized vector previously obtained.

Figure 1| Synthetized insert. The prefix and suffix are located in the upstream (5') and downstream (3') regions of this sequence, respectively flanked by HindIII and SacI cutting sequences. At the ends of the insert there are 6 base pairs (called "landing bases") necessary to allow the correct binding and recognition of the restriction sites, located just after the 6 base pairs, by the restriction enzymes with which the sequence was later digested. The landing bases are also useful for minimizing errors in important regions during the synthesis. Internally, the standardized promoter iGEM BBa_J23119 is also copied.

The resulting plasmid has a size of 5966 bp and is named pSGAb iGEM (BBa_K4727000). The Prefix and Suffix elements also allow for the replacement of the reporter gene or a gene of interest through an enzymatic digestion reaction followed by a ligation reaction. In our case, the gene of interest, and its relative expression cassette, corresponds to the dCas9 and the sgRNAs sequences.

Cloning a RFP reporter system in BBa_K4727000

The pSGAb iGEM plasmid is able to replicate in A. baumannii and K. pneumoniae, as well as in E. coli. The former two are non-model organisms for which electroporation protocols have been optimized. It was therefore necessary to clone an expression system for a reporter gene, and we chose the rfp gene.

As already mentioned in the experiments section of this Wiki, insertions in the plasmid were attempted of both the iGEM BBa_I13507 cassette, creating a back insert (resulting in the expression of the RFP under the J23119 promoter), and the BBa_I13521 cassette, first again as a back insert, then by total replacement with BBa_J23119. The initial results led to considerations and conclusions well described in the Engineering section entitled "A matter of promoters". In short, the plasmids pSGAb_I13507 and pSGAb_I13521 were successfully obtained.

pSGAb_I13507 (Figure 2), which has a size of 6835 bp, was mainly used in the transformation of A. baumannii and K. pneumoniae.

Figure 2| Reporter plasmid pSGAb_I13507 in A. baumannii and K. pneumoniae.

After the electroporation of the two non-model strains, in both cases red colonies were observed. Furthermore, plate readers have been performed to evaluate the expression of the RFP. This result was essential, given that the BBa_I13507 cassette, as well as BBa_I13521, contains the Ribosome Binding Site BBa_B0034 sequence and the terminator BBa_B0015: we found out that these genetic parts could function in A. baumannii, but there was no information about K. pneumoniae in the literature. Therefore, with these experiments we were able to confirm the potential use of the expression cassette for the dCas9 contained in the plasmid J116-dCas9_m_3k3 (PCR mutagenesis results of this section).

Later, we measured the transformation efficiency in the two strains and mutation rate of the rfp gene in A. baumannii (Results, Bacterial transformation).

The synthesis rate of the J23119 promoter was also assessed in both the strains (Contributions, BBa_J23119 expressing proteins in A. baumannii, BBa_J23119 expressing proteins in K. pneumoniae), as well as the synthesis rate of pTet following the transformation of A. baumannii with pSGAb_I13521 (Contributions, pTet (BBa_R0040) expressing proteins in A. baumannii).

The cloning of the dCas9 cassette and BBa_K4727009

As previously mentioned, the initial intention was to clone the expression cassette of the dCas9 (whose CDS was standardized by creating BBa_K4727008), present in the J116-dCas9_m_3k3 plasmid, inside the pSGAb backbone we obtained. Unfortunately, several cloning attempts were unsuccessful, starting both with BBa_K4727000 and pSGAb_I13507, and performing both back and front insertions. The process of result interpretation and the subsequent operations performed are explained in detail in the Engineering section, "Dealing with a cloning problem". In conclusion, the original ori of BBa_K4727000 was modified, resulting in the plasmid BBa_K4727009.

BBa_K4727001: a standard backbone for P. aeruginosa

First Gibson assembly: pUCP20_RFP_rrnbT12

To achieve the final plasmid, we had to perform a first assembly to build a non-BioBrick plasmid expressing RFP with suitable parts for P. aeruginosa. This plasmid was called pUCP20_RFP_rrnbT12 and, as the name suggests, was composed by:

  • pUCP20 backbone (Ori pMB1, AmpR cassette, pLac promoter)
  • BBa_E1010 gene (RFP), controlled by the pLac promoter, from BBa_I13507
  • rrnbT1 and rrnbT2 terminators, from pMF230

In the pUCP20 backbone the RBS is not annotated; however, after transformation in E. coli the RFP was expressed, so it must be present.

In the earlier stage of the activities, we thought that we would be able to use ampicillin as a selection marker for the transformation of P. aeruginosa; we developed this non-BioBrick plasmid to test the expression in this bacteria, so we could proceed both with the optimization of the electroporation protocol and with the plasmid development. However, during the first attempts of transformation, we noticed that the tolerance of the bacteria to this antibiotic was canceling our efforts. We then realized that we needed to change the selection marker gene from ampicillin to chloramphenicol, which is the only antibiotic allowed in the iGEM rules to which the strain P. aeruginosa ATCC 10145 is sensible.

Standardization

To standardize the plasmid, we added prefix and suffix sequences between the promoter pLac (from pUCP20 PilB) and the terminators rrnbT1 and rrnbT2 (from pMF230). The plasmid in this phase was called pUCP20 standard. During this mutagenesis, the RFP gene was lost, so it will be necessary to clone it using digestion ligation, according to the BioBrick standard. The resulting plasmid has a selection marker for ampicillin and the prefix and suffix sites downstream of the promoter and upstream of the terminator, respectively.

To obtain the final plasmid, the next steps are to modify the selection marker (from ampicillin resistance to chloramphenicol resistance) and to clone the BBa_I13507 cassette inside the prefix and suffix.

Before proceeding with cloning it is necessary to consider whether to perform the cloning of the reporter gene first or before the resistance modification. By inserting the RFP gene first, you get a construct that provides qualitative information about transformation efficiency; in fact, the colonies transformed with plasmid are visibly red. However, the strain used in the laboratory has resistance to ampicillin so processing is not possible.

Another aspect to consider, however, is the advantage of producing the stock of a backbone functioning in P. aeruginosa: by first modifying the resistance box, you get the plasmid with prefix and suffix intact, without any cloned gene inside them; this version of plasmid is very useful because it allows you to have a backbone ready to clone any gene, from another plasmid BioBrick, inside of it.

In the specific case of this project, in this first phase an RFP will be cloned, later the gene for dCas9 will be inserted. Given these advantages, it is believed that the most convenient solution is to first run the Gibson assembly for resistance change and only later clone the RFP gene within prefix and suffix.

Second Gibson assembly: from AmpR to ChlR

To obtain the final plasmid, it is necessary to change the selection marker gene from ampicillin to chloramphenicol. We performed a second Gibson assembly using fragments coming from pUCP20 standard and the standard backbone pSB4C5, from which we got the gene for chloramphenicol acetyl-transferase for the chloramphenicol resistance.

In both the Gibson assemblies, we found that, as mentioned in the manufacturer’s instructions, we needed to extend the reaction time up to 1h to make it work.

At this point we obtained the backbone pPAC, registered as BBa_K4727001.

Electroporation protocol optimization for Acinetobacter baumannii

The electroporation procedure has been optimized considering the following protocols: Biswas e Rather 2019 [1] e Jacobs et al., 2014 [2].

The ratio of LB volume to inoculum tube capacity has been kept equal to 0,3, while the ratio of inoculum volume to flask capacity used for the second inoculum has been kept equal to 0,1. These low ratios are required since A. baumanii is an aerobic organism that needs a large amount of air to properly grow (Yildirim et al. 2016 [3]). For the same reason, incubator shaking must be kept at 220 rpm.

Since the procedure calls for two inocula, OD600 turns out to be very high and for this reason the pellet appears to have a considerable size and is tightly adherent to the vial side. Besides having a good quality pellet, the main reason why two inocula are made is to reach the beginning of the stationary growth phase, which is the best time to electroporate A. baumanii, as demonstrated by Biswas e Rather 2019 e Yildirim et al., 2016 researches.

The electroporation parameters used (1800 V, 100 Ω, 25 µF) were chosen after previously testing those reported in the Jacobs et al., 2014 protocol (1700 V, 100 Ω, 25 µF), which, however, had yielded negative results.

The plasmid chosen to test electroporation was pSGAb-I13507– coming from the cloning of gene reporter RFP expression cassette BBa_I13507 in BBaK4727000 backbone, resulting from the pSGAb-km plasmid optimization (Results section, BBa_K4727000).

Using the protocol outlined in the experiments section (Electroporation Acinetobacter baumannii), we achieved the growth of both red and white colonies. Therefore, we opted to select four colonies, including three white colonies and one red colony (sample 2.2A in Figure 1,a).

To confirm the electroporation results and the presence of BBaK4727000 carrying the RFP gene within the bacteria, we submitted the samples for sequencing using the FW_seq_pSGAb and RV_seq_pSGAb primers (experiments section of this Wiki, pSGAb). The sequencing results confirmed the presence of the RFP gene in all four samples analyzed. Additionally, we carried out a MALDI-TOF test (courtesy of the Complex Microbiology Operational Unit of Padua Hospital) to ensure the colonies were indeed A. baumannii. The results were all positive.

Moreover, despite Jacobs et al. 2014 reporting a decreased transformation efficiency of samples stored at -80 oC, the electroporation tests were successfully performed and no significant loss of efficacy was detected.

Acinetobacter baumannii transformation and pSGAb_I13507 stability

To comprehend the cause behind the presence of white colonies, we performed two experiments: serial colony count (experiments section, Mutation Rate assessment in A. baumannii) and microplate reader analysis.

Starting from an LB+kanamycin plate containing electroporated A. baumannii colonies, in which the plasmid pSGAb_I13507 was inserted, a single red colony was taken and inoculated into a selective medium. The next day, the inoculum was plated with 1:1,000 and 1:10,000 dilutions. Once the colonies had grown, they were counted. A total of 5 counts were performed for each of the two dilutions, for a total of 10 plates. The initial hypothesis is that the white colonies are due to a mutation in the RFP expression cassette, this mutation inhibits the expression of the gene by going to lighten the metabolic load of the mutated bacteria, making this mutation favorable.

A. ba. colony counts

Table 1: Data obtained from the colony counts of A. baumannii electroporated with pSGAb_I13507. The ratio was calculated by dividing the number of red colonies with the total number of colonies (red plus white).

As can be seen in the table, the ratios of white colony occurrence seem to always stay around 80%. However, the ratios of white to red colonies in test N°3 have a red colony value of 0, despite this the inoculum grown from the selection of one of these white colonies was red, and it can also be seen in the table that the colonies grown in test N°4 have a ratio of red to white colonies similar to tests N°1 (Image 1), N°2 and N°5.

A. ba. electroporation

Figure 1|Colonies of A. baumannii electroporated with pSGAb_I13507. (a) Smear of colony 2.2A found to be visibly red in the first electroporation. (b) Colony count, test plate 1.

The fluorescence analyses of the samples used to conduct the RFP expression readout at the plate reader are explained in detail in the Contribution section, BBa_J23119 expressing proteins in A. baumannii. Briefly, four previously mentioned colonies named 2.2A, 1.1A, 1.2A and 1.2B electroporated with pSGAb_I13507 were compared with E. coli TOP10 strain expressing the RFP gene under the control of both the relatively weak promoter BBa_J23101 and the strong promoter BBa_R0040 (pTet). In particular, colony 2.2A distinctly displayed heightened expression. Upon subjecting the expression cassettes to Sanger sequencing, we found out a mutation within the RFP coding sequence in 1.1A, 1.2A and 1.2B. This causes a reduction in emission intensity.

Here, Turbidimetric analysis are particularly discussed (Figure 2). These show that A. baumannii WT has a higher growth than samples 2.2A, 1.2A and 1.1A but is comparable to the growth of sample 1.2B; while with regard to sample 2.2A a lower growth was to be expected given the higher expression of RFP which has a certain metabolic weight, for samples 1.2A and 1.1A on the other hand, a trend more similar to sample 1.2B was expected. Furthermore, it can be seen that the growth of E. coli TOP 10 pTet-RFP is greater than that of sample 2.2A despite the fact that fluorescence analysis shows a greater presence of RFP protein and the growth of A. baumannii WT is greater than the growth of E. coli WT. These data could indicate how RFP protein expression is particularly onerous for the A. baumannii ATCC 19606 strain compared to a model strain such as E. coli TOP 10.

A. ba. growth

Figure 2| Turbidimetric analysis of the samples

If the initial hypothesis had been correct, subsequent a to test N°3, the colonies should no longer exhibit a red phenotype, whereas both the inoculum and subsequent tests show visibly red colonies. Furthermore, comparing the growth curves, all colonies that do not visibly express RFP should have a metabolic advantage and thus more growth, which is, however, only visible in one of the samples.

The hypothesis that was therefore subsequently formulated is that within the same strain of A. baumannii ATCC19606 there may be some variability in the expression of RFP protein (in this case) and that therefore it is not always visible: since A. baumannii ATCC19606 is not a model bacterium it is believed that intrinsic biological variability is present.

In the case of sample 1.2B which showed more growth than the other transformed samples the variation could be due to less mRNA formation which would explain both the little fluorescence and the more growth; in the case of the colonies showing little fluorescence and less growth the expected mRNA levels would be higher indicating that the problem could be incomplete translation of the RFP protein or the formation of a protein that is not fully functional also due to too much metabolic load required.

Transformation of Klebsiella pneumoniae

The protocol for making Klebsiella pneumoniae cells competent from Wang et al., 2018 [4] presented several critical points. Specifically, after resuspending the bacteria in 15 ml of 10% (v/v) glycerol and centrifuging them, we did not achieve complete separation of the pellet from glycerol, leading to the loss of bacterial sediment during the necessary inversion step to remove the supernatant. To overcome this issue, modifications were made to the glycerol removal method by adopting a removal approach using a pipette tip. However, due to the formation of an aggregate between concentrated glycerol and the bacterial pellet, it was impossible to completely remove glycerol from Klebsiella pneumoniae cells. The presence of glycerol negatively affected the subsequent electroporation phase.

To remove glycerol, two additional centrifugations at 13,000 rpm for 3 minutes were performed. Even in this case, the result obtained was a completely homogeneous solution in which bacteria and glycerol were uniformly mixed, probably due to their similar densities.

Based on the results obtained, the protocol was further modified by reducing the centrifugation speed to 8,000 rpm for 15 minutes to allow the separation of the bacterial pellet from the supernatant without affecting the separation of glycerol from water. But even this attempt was unsuccessful, resulting in glycerol concentration at the bottom of the tube and no separation of the bacterial pellet. In light of the results obtained, we proceeded with the transformation protocol previously used for Acinetobacter baumannii (Experimental section, Electroporation Acinetobacter baumannii).

However, it is important to note that the protocol optimized for Acinetobacter baumannii requires stationary-phase growth with a much higher OD compared to that reported in Klebsiella pneumoniae electroporation protocols, which call for a mid-exponential growth phase. Additionally, the pellet formed was not stable due to Klebsiella pneumoniae being a mucoid bacteria, and it is believed that its extracellular matrix interacted with glycerol, resulting in a homogeneous solution.

To overcome this issue, we attempted a new protocol (Experimental section, Klebsiella pneumoniae electroporation with citric acid) based on the use of citric acid as a solution to remove the bacterial extracellular matrix. The washing steps confirmed this hypothesis, as the pellet produced from bacterial cultures grown in LB with the addition of citric acid adhered more to the tube walls, allowing for effective removal of glycerol without affecting the pellet itself. Furthermore, cultures were maintained in a mid-exponential growth phase to achieve an OD600 between 0.5-0.6.

Electroporation with pSGAb_I13507 (the same of A. baumannii) was performed in 0.1 cm cuvettes, using the following settings: 1.8 kV, 200 Ω, and 25 uF. After electroporation, bacteria were transferred to an Eppendorf tube containing 1 ml of preheated SOC medium and incubated for one hour and thirty minutes.

Subsequently, 100 uL of concentrated and 1:100 diluted bacteria were plated on LB plates containing kanamycin to confirm the results of electroporation and the presence of pSGAb, which carries the RFP gene inside the bacteria. The grown colonies were then subjected to sequencing using the FW_seq_pSGAb and RV_seq_pSGAb primers (experiments section of this Wiki, pSGAb). The sequencing results successfully confirmed the presence of the RFP gene and hence the success of the transformation.

Alternative transformation protocol

In the early stages of our wet lab activity, we had to look for protocols for the transformation of non standard species (namely, A. baumannii, K. pneumoniae and P. aeruginosa). For these bacteria, we found that the most common method was electroporation, but unfortunately we encountered some difficulties in the use of this method. Therefore, while trying to optimize the electroporation, we looked for an alternative way for transformation.

We found a new technique that we never heard of before, based on the Yoshida effect. Described for the first time in 2007 by N. Yoshida [5], this method relies on DNA uptake mediated by friction forces: a plasmid-cell suspension is mixed with mineral nanofibers, then dropped onto the surface of an elastic body, such as an agarose plate, and treated physically by sliding a polystyrene streak bar over the elastic body to create friction.

The sliding friction forces, arising between the surface of the agar and the stir stick when bacteria are spread, result in penetration of bacterial cells. This leads to inoculation of the transforming DNA, which is adsorbed to the mineral nanofibers. The first protocol assessed for E. coli uses chrysotile nanofibers; however, this mineral is known to possibly cause health damage [6], so we found two optimized protocols that use sepiolite, which is a naturally occurring magnesium silicate that has been mentioned by Yoshida and colleagues as an alternative source of nanofibers [7].

Principle and procedure of the nanopiercing method. (a) Transforming DNA adsorbs to sepiolite fibers. (b) Transforming DNA, sepiolite suspension and bacteria are mixed and plated on a selective agar medium; sliding friction forces developing during plating cause piercing of bacterial cells with sepiolite fibers. (c) Inside the bacterial cell, transforming DNA possibly is competitively displaced from sepiolite fibers by cellular nucleic acids.

In particular we found one protocol for E. coli and A. baumannii [7] and another one for P. aeruginosa [8]. This method seems to be effective with Gram negative bacteria. It is a rapid and cost-effective method, it doesn’t need the preparation of competent cells and it’s characterized for three out of four of our bacterial targets.

However, it has some disadvantages. The main problem involves the antibiotics that we use as selection markers. As explained in the protocol, antibiotics that interfere with protein synthesis, such as kanamycin and chloramphenicol, impede resistance marker production and therefore cannot be present at the time of transformation. For this reason, we first tried to transform E. coli with BBa_I13521, which expresses RFP as a reporter gene and has ampicillin resistance as marker of selection. This transformation was successful, but some of our bacterial strain (i.e. K. pneumoniae and P. aeruginosa) were ampicillin resistant; so, an optimization was needed.

Our efforts focused on optimizing the protocol in order to be able to transform bacteria using kanamycin as marker of selection. To overcome the problem, the protocol suggests transforming on plates with no antibiotic, then incubating plates 1 or 2 h at 37 °C and then adding the appropriate amount of antibiotic at a volume of no less than 200 μl to ensure the homogeneous spread of the antibiotic over the surface. We tried this method, adding to non-selective plates about 200-400 ul (in different attempts) of kanamycin at the appropriate concentration (25 ug/ml), but with no success.

Unfortunately, we didn’t manage to optimize the protocol for transformation with kanamycin or chloramphenicol, which would have been useful to deal with non-standard bacterial strains such as the ESKAPE. Anyway, it is a very efficient and fast method of transformation of E. coli with plasmids with ampicillin resistance as marker of selection.

References

  1. Biswas, Indranil e Philip N. Rather, cur. (2019). Acinetobacter baumannii: Methods and Protocols. en. Vol. 1946. Methods in Molecular Biology. New York, NY: Springer New York. ISBN: 978-1-4939-9117-4 978-1-4939-9118-1. DOI: 10. 1007/978-1-4939-9118-1. URL: http://link.springer.com/10.1007/978-1-4939-9118-1
  2. Jacobs, Anna C. et al. (nov. 2014). ªGenetic Manipulation of Acinetobacter baumanniiº. en. In: Current Protocols in Microbiology 35.1. ISSN: 1934-8525, 1934- 8533. DOI: 10.1002/9780471729259.mc06g02s35. URL: https://onlinelibrary.wiley.com/doi/10.1002/9780471729259.c06g02s35.
  3. Yildirim, Suleyman et al. (feb. 2016). ªEvaluation of Parameters for High Efficiency Transformation of Acinetobacter baumanniiº. en. In: Scientific Reports 6.1, p. 22110. ISSN: 2045-2322. DOI: 10.1038/srep22110. URL: https://www.nature.com/articles/srep22110.
  4. Wang, Y., Wang, S., Chen, W., Song, L., Zhang, Y., Shen, Z., Yu, F., Li, M., & Ji, Q. (2018). CRISPR-Cas9 and CRISPR-Assisted Cytidine Deaminase Enable Precise and Efficient Genome Editing in Klebsiella pneumoniae. Applied and Environmental Microbiology, 84(23), e01834-18. https://doi.org/10.1128/AEM.01834-18.
  5. Yoshida N, Nakajima-Kambe T, Matsuki K, Shigeno T. Novel plasmid transformation method mediated by chrysotile, sliding friction, and elastic body exposure. Anal Chem Insights. 2007 Feb 14;2:9-15. PMID: 19662172; PMCID: PMC2716810.
  6. Philip J. LANDRIGAN, William J. NICHOLSON, Yasunosuke SUZUKI, Joseph LADOU, The Hazards of Chrysotile Asbestos: A Critical Review, Industrial Health, 1999, Volume 37, Issue 3, Pages 271-280, Released on J-STAGE March 29, 2007, Online ISSN 1880-8026, Print ISSN 0019-8366, https://doi.org/10.2486/indhealth.37.271.
  7. Gottfried Wilharm, Daniela Lepka, Franziska Faber, Julia Hofmann, Tobias Kerrinnes, Evelyn Skiebe. A simple and rapid method of bacterial transformation, Journal of Microbiological Methods, Volume 80, Issue 2, 2010, Pages 215-216,ISSN 0167-7012, https://doi.org/10.1016/j.mimet.2009.12.002.
  8. Jerónimo Rodríguez-Beltrán, Hamouda Elabed, Kamel Gaddour, Jesús Blázquez, Alexandro Rodríguez-Rojas. Simple DNA transformation in Pseudomonas based on the Yoshida effect, Journal of Microbiological Methods, Volume 89, Issue 2, 2012, Pages 95-98, ISSN 0167-7012, https://doi.org/10.1016/j.mimet.2012.02.013.