Here, we describe a simple and reproducible protocol of mouse model of infection to evaluate the attenuation of the genetically modified strains of Pseudomonas aeruginosa in comparison to the United States Food and Drug Administration (FDA)-approved Escherichia coli for commercial applications.
Microorganisms are genetically versatile and diverse and have become a major source of many commercial products and biopharmaceuticals. Though some of these products are naturally produced by the organisms, other products require genetic engineering of the organism to increase the yields of production. Avirulent strains of Escherichia coli have traditionally been the preferred bacterial species for producing biopharmaceuticals; however, some products are difficult for E. coli to produce. Thus, avirulent strains of other bacterial species could provide useful alternatives for production of some commercial products. Pseudomonas aeruginosa is a common and well-studied Gram-negative bacterium that could provide a suitable alternative to E. coli. However, P. aeruginosa is an opportunistic human pathogen. Here, we detail a procedure that can be used to generate nonpathogenic strains of P. aeruginosa through sequential genomic deletions using the pEX100T-NotI plasmid. The main advantage of this method is to produce a marker-free strain. This method may be used to generate highly attenuated P. aeruginosa strains for the production of commercial products, or to design strains for other specific uses. We also describe a simple and reproducible mouse model of bacterial systemic infection via intraperitoneal injection of validated test strains to test the attenuation of the genetically engineered strain in comparison to the FDA-approved BL21 strain of E. coli.
Pseudomonas aeruginosa is an opportunistic bacterial pathogen that can cause life-threatening diseases in humans, especially in the immunocompromised. The pathogenicity of P. aeruginosa is due to the expression of many virulence factors, including proteases and lipopolysaccharide, as well as its ability to form a protective biofilm1. Because of its ability to produce virulence factors and cause disease in humans, using P. aeruginosa to make commercial products presents safety concerns. Nonpathogenic strains of E. coli have traditionally been used to bioengineer medical and commercial products for human use. However, some products are difficult for E. coli to make, and many are packaged in inclusion bodies, making extraction laborious. Engineered bacterial strains with the ability to make and secrete specific products is highly desirable, as secretion would likely increase yield and ease purification processes. Thus, nonpathogenic strains of other species of bacteria (e.g., species that utilize more secretion pathways) may provide useful alternatives to E. coli. We recently reported the development of a strain of P. aeruginosa, PGN5, in which the pathogenicity and toxicity of the organism is highly attenuated2. Importantly, this strain still produces large quantities of the polysaccharide alginate, a commercially interesting component of the P. aeruginosa biofilm.
The PGN5 strain was generated using a two-step allelic exchange procedure with the pEX100T-NotI plasmid to sequentially delete five genes (toxA, plcH, phzM, wapR, aroA) known to contribute to the pathogenicity of the organism. pEX100T-NotI was generated by changing the SmaI to a NotI restriction enzyme recognition site within the multiple cloning site of the plasmid pEX100T, which was developed in Herbert Schweizer's lab3,4. The recognition site for the restriction enzyme NotI is a rarer DNA sequence compared to SmaI and less likely to be present in sequences being cloned, thus it is more convenient for cloning purposes. The plasmid carries genes that allow for selection, including the bla gene, which encodes ß-lactamase and confers resistance to carbenicillin, and the B. subtilissacB gene, which confers sensitivity to sucrose (Figure 1A). The plasmid also carries an origin of replication (ori) compatible with E. coli, and an origin of transfer (oriT) that allows for plasmid transfer from E. coli à Pseudomonas species via conjugation. However, the plasmid lacks an origin of replication compatible with Pseudomonas, and thus cannot replicate within Pseudomonas species (i.e., it is a Pseudomonas suicide vector). These characteristics make pEX100T-NotI ideal for targeting genetic deletions from the Pseudomonas chromosome. Plasmid cloning steps are carried out using E. coli and the resultant plasmid is transferred to Pseudomonas by transformation or conjugation. Then, through homologous recombination events and selective steps, the targeted in-frame deletion is generated, marker-free. This method of sequentially deleting genomic regions from the chromosome of P. aeruginosa could be used to generate highly attenuated Pseudomonas strains, like PGN5, or to design strains for other specific uses (e.g., strains deficient in endonucleases for plasmid propagation or strains deficient in proteases for production of proteins of interest).
The overall virulence of strains of bacteria is affected by growth conditions and phases, during which mutations occur frequently. Therefore, measuring the safety of genetically-engineered strains can be challenging. To evaluate bacterial isolates for systemic virulence, we adapted a previously published protocol of infection by intraperitoneal injection of C57BL/6 mice5. We modified this procedure to use frozen bacterial stocks for injection, which allowed for precise dosing and easy validation of the strains used. In this model, the E. coli strain BL21, which has been FDA-approved for production of biopharmaceuticals, was used as a control safety standard for determining the relative pathogenesis of the strain6,7,8. The main advantage to using this method is that it is reproducible and minimizes sources of variation, as infecting strains are validated for bacterial cell number, phenotype, and genetic markers both before and after infection. With these controlled steps, the number of animals required is reduced. In this model, P. aeruginosa strains that result in C57BL/6 murine mortality rates equal to or less than E. coli BL21 when injected intraperitoneally may be considered attenuated. This simple mouse model of infection may also be used to assess the attenuated pathogenicity of genetically engineered strains from other species using the FDA-approved E. coli strain as the reference. Steps 1-7 detail the generation of sequential genomic deletions in P. aeruginosa (Figure 1) and steps 8-12 detail the use of a mouse model to test the pathogenicity of P. aeruginosa strains.
Before beginning animal experiments, the protocol to be used must be approved by the Institutional Animal Care and Use Committee (IACUC). Approval for the protocol described was obtained through the IACUC at Marshall University (Huntington, WV, USA).
1. Plasmid Design
2. Plasmid Preparation
3. E. coli Transformation
4. Bacterial Strain Preparation and Triparental Conjugation
5. Detection of Single-crossover Recombinants of P. aeruginosa
6. Detection of Double-crossover Recombinants of P. aeruginosa
7. Gene Deletion Confirmation via Colony PCR
8. Preparation of Bacterial Strain for Animal Testing
9. Validation of Growth and Strain of Stocks Stored for Animal Testing.
10. Inoculation of Animals with Bacterial Strains by Injection
11. Statistical Analysis of Animal Mortality
12. Visualization of the Infection with Bioluminescence
As shown in Figure 2, the targeted genomic deletion can be confirmed using colony PCR with specific primers that amplify the region of interest. Colonies that carry a genomic deletion will yield a shorter PCR band size in comparison to wild-type colonies. A PCR-screen of 10-12 colonies is usually sufficient to detect at least one colony that carries the targeted deletion. If no deletions are detected after multiple rounds of screens, repeat the procedure beginning with the conjugation. If the deletion still fails, the plasmid insert may need to be confirmed through sequencing, redesigned, or the deletion may be lethal. Upon the verification of a gene deletion via PCR, confirm the deletion through sequencing. The resulting strain may be subjected to the procedure repeatedly to generate sequential genomic modifications.
As shown in Figure 3, mortality associated with intraperitoneal injection of the attenuated strain of P. aeruginosa PGN5 (+mucE) was 0%, which was equivalent to mortality observed with E. coli BL21. On the other hand, intraperitoneal injection of the parent strain (VE2) was fatal to 80% of mice. These results were obtained with extensive steps to validate the strains injected. While the exact cause of death in these mice is unknown, it can at least in part be attributed to the expression of virulence factors in the parent strain that were deleted from the attenuated PGN5 strain. Differences in the infection progression was tracked using bioluminescence-marked parent and attenuated strains. The attenuated strain remained localized at the site of injection until bioluminescence faded (Figure 4). The clearance of the infection most likely coincided with the fading of the bioluminescence. Bioluminescence was not detected 24 h after injection and mice lived for weeks following injection until sacrificed, with no adverse effects observed.
Figure 1: Generating gene deletions in P. aeruginosa with pEX100T-NotI. (A) Map of the pEX100T-NotI plasmid. (B) Generation of a construct composed of regions directly upstream (yellow) and downstream (blue) of the region of interest (ROI), flanked with NotI restriction enzyme recognition sites. First, PCR-amplify upstream and downstream regions independently with specific primers that add 5' NotI digestion sites (e.g., NotI-aroA F CGCGGCCGCTGAAGGTCCTGGGCTCCTATCCGAAAGCGGTGCTCT and NotI-aroA R GCGGCCGCAGTTGGGTTGTTCTGCGATGGCGCCAGGCA) and 3' overlapping homologous regions as shown (e.g., aroA-crossover F CTCCAGGCGCTGGGCAAGGTGCTGGCGCATGACTGAGGTCACGCCGGTCGCCGTGGAGAACA and aroA-crossover R TGTTCTCCACGGCGACCGGCGTGACCTCAGTCATGCGCCAGCACCTTGCCCAGCGCCTGGAG. Then, use PCR withNotI-containing primers to join the upstream and downstream products generated in the first PCR reaction. (C) The pEX100T-NotI plasmid, armed and ready. Ligate the NotI-digested cross-over PCR product into the NotI-digested plasmid. (D) Flow diagram of the process to delete genomic regions from the P. aeruginosa chromosome using the pEX100T-NotI plasmid. After the desired deletion has been confirmed and purified, the resultant strain can be taken through the procedure repeatedly to delete other genomic regions from the chromosome. When the desired strain is obtained, sequence the whole genome to confirm deletions and other changes to the chromosome. The pathogenicity of the strain can then be tested in mice using the procedure outlined in Part II of the Protocol. Please click here to view a larger version of this figure.
Figure 2: Gel electrophoresis of colony PCR products from a screen for aroA deletion to generate the attenuated P. aeruginosa strain, PGN5. Colony PCR products run in lanes 2-5 and 8-11 indicate colonies with wild-type aroA. Colony PCR products run in lanes 6 and 7 carry the aroA gene deletion, indicated by the smaller PCR product (yellow asterisks). Primers used specifically amplified the genomic region containing the aroA gene: aroA-F: GCGAACGCCAACAGCCGATAAAGC, and aroA-R: ATCTGGCTCGCGATGCCGGTCC. Expected PCR product size in wild-type colonies was 2,548 nucleotides (nt). Expected PCR product size in colonies with aroA deletion was 307 nt. A DNA ladder was run in lanes 1 and 12. Please click here to view a larger version of this figure.
Figure 3: Overall mortality of mice injected with pathogenic P. aeruginosa strain (VE2), attenuated P. aeruginosa strain (PGN5+mucE), and FDA control E. coli strain (BL21). Only mice injected with pathogenic parent strain exhibited mortality at 80%. Attenuated P. aeruginosa strain and FDA control E. coli strain exhibited 0% mortality. Please click here to view a larger version of this figure.
Figure 4: Image of mouse 3 h post-injection of attenuated strain of P. aeruginosa PGN5+mucE carrying a bioluminescent marker. The bioluminescent bacteria were detectable until 18-24 h following injection. During this period, the bioluminescence remained at the site of injection indicating the bacteria stayed localized to injection site. This mouse fully recovered with no adverse effects. Please click here to view a larger version of this figure.
The pEX100T-Not1 plasmid is an efficient mediator of sequential genomic deletions that are marker-free and in-frame. When engineering bacterial strains for attenuated virulence, deletion of entire gene sequences rather than generating point mutations decreases the likelihood of reversion to a virulent phenotype. Additionally, each pathogenicity gene deletion attenuates the pathogen further, reinforcing the stability of the attenuation.
This method can also be used to generate genomic modifications other than deletions, such as point mutations and insertions, simply by modifying the design of the plasmid insert. These types of modifications may be more useful than entire gene deletions for engineering bacteria with modified metabolism, for example. Sequential genomic modification has significant potential for generating designer bacterial strains to suit specific purposes in research and industry. Other methods of generating desired marker-free genomic modifications in bacteria have been described15,16,17,18. As with all genome-editing methods, attempted modifications to essential genomic regions may be lethal, and thus unsuccessful. In these cases, identification of different genetic modifications or other candidate genes is required to generate the bacterial strain of interest.
Given the numerous replication events and passages of each colony throughout this protocol, unintended changes to the genome will occur to the generated strain. The exact genomic changes can be identified through whole-genome sequencing. However, the impact of these changes is harder to determine. When engineering bacteria for a specific purpose, genomic changes that do not negatively affect the growth of the organism or the targeted pathway(s) are tolerable. Depending on the strain being generated, it may be possible to identify a "readout" to ensure that the strain is still useful for its intended purpose. For example, with PGN5, the goal was to create an attenuated strain that retained the ability to produce large amounts of alginate. After deletion of five pathogenicity genes, the amount and composition of alginate produced by PGN5 was measured and determined to be comparable to other alginate-producing strains. Thus, alginate production was unaffected by the five gene deletions, nor by the unintended genomic changes that occurred during the development of PGN5.
A model of intraperitoneal mouse injection was used to determine whether an engineered strain was attenuated compared to the parent strain and E. coli BL21, a strain approved by the FDA for production of biopharmaceuticals. The most important steps taken during this animal testing procedure were preparation and validation of frozen bacterial stocks. Preparation and use of frozen bacterial cultures to inject mice is preferable to using continuous culture, as it reduces the number mutations that naturally occur in bacterial populations19. Additionally, frozen cultures should remain viable for years. Viable plate counts showed no significant difference between the CFU/mL directly after stocks were prepared and three months after preparation. The use of multiple validation steps throughout this procedure ensured that the method was reproducible, and the results were not skewed by contaminating bacteria. Additionally, with the number of precautionary steps taken to ensure reproducibility, fewer animals were needed. Using a bacterial strain that is FDA-approved for biopharmaceutical production as the control (such as E. coli strain BL21), this method could be used to test the attenuation of other genetically engineered strains of P. aeruginosa, or other species of bacteria.
Using bioluminescence as a marker provides additional validation of the bacterial strains injected, as the marker can be visualized at the injection site. Insertion of the bioluminescence marker into the bacterial chromosome is required for bioluminescence imaging but may not be possible if working with incompatible strains/species. However, marking strains with bioluminescence is not required to test for attenuation. The strains tested in this study were marked with bioluminescence, which allowed for visualization of localization differences between strains throughout the course of the infection. We observed that the pathogenic strain disseminated through the body of the mouse, but the non-pathogenic strain remained at the site of injection. While this experiment only tested two very closely related strains of P. aeruginosa, it suggests that bacterial dissemination is linked to virulence, at least in P. aeruginosa. Thus, this procedure of labeling with bioluminescence to visualize the progression of the infection could be used in the future to quickly evaluate the attenuation of engineered strains of bacteria.
The authors have nothing to disclose.
This work was supported by the National Institutes of Health (NIH) grants R44GM113545 and P20GM103434.
0.2 mL tubes with flat caps | ThermoScientific | AB-0620 | via Fisher Scientific |
1 mL Syringe | BD | 22-253-260 | via Fisher Scientific |
1.5 mL disposable polystyrene cuvette | Fisher Scientific | 14955127 | |
1.5 mL Microcentrifuge Tubes | Fisher Scientific | 05-408-129 | |
2.0 mL Cryogenic Vials | Corning | 430659 | via Fisher Scientific |
27G needle | BD | 14-821-13B | via Fisher Scientific |
50 mL tubes | Fisher Scientific | 05-539-13 | via Fisher Scientific |
Accu block Digital Dry Bath | Labnet | NC0205808 | via Fisher Scientific |
Benchtop Centrifuge 5804R | Eppendorf | 04-987-372 | via Fisher Scientific |
Benchtop Microcentrifuge | Sorvall | 75-003-287 | via Fisher Scientific |
Cabinet Incubator | VWR | 1540 | |
Carbenicillin disodium salt | Fisher Scientific | BP2648250 | |
Culture Test Tube, Polystyrene | Fisher Scientific | 14-956-6D | via Fisher Scientific |
Diposable Inoculation Loops | Fisher Scientific | 22-363-597 | |
Dneasy UltraClean Microbial Kit (50) | Qiagen | 12224-50 | or preferred method/vendor |
E.Z.N.A. Cycle Pure Kit (50) | Omega bio-tek | D6493-01 | or preferred method/vendor |
EcoRI-HF, restriction endonuclease | New England BioLabs | R3101L | |
Electroporation Cuvettes | Bulldog Bio | NC0492929 | via Fisher Scientific |
FastLink II DNA Ligation Kit | Epicentre Technologies | LK6201H | via Fisher Scientific |
Gentamycin Sulfate | Fisher Scientific | BP918-1 | |
Glycerol | Fisher Scientific | BP229-4 | |
GoTaq G2 Colorless Master Mix | Promega | M7833 | via Fisher Scientific |
Isothesia Isoflurane | Henry Schein Animal Health | 29405 | |
IVIS Lumina XRMS Series III In Vivo Imaging System | Perkins and Elmer | CLS136340 | |
Kanamycin monosulfate | Fisher Scientific | BP906-5 | |
LE agarose | Genemate | 3120-500 | via Fisher Scientific |
Luria Broth | Difco | 240230 | via Fisher Scientific |
MicroPulser Electroporator | BioRad | 1652100 | |
Noble agar, ultrapure | Affymetris/USB | AAJ10907A1 | via Fisher Scientific |
NotI-HF, restriction endonuclease | New England BioLabs | R3189 | |
One Shot TOP10 Electrocomp E. coli | Invitrogen | C404052 | via Fisher Scientific |
Phosphate buffered saline powder | Sigma | P3813-10PAK | Sigma-Aldrich |
Prism 7 | GraphPad | https://www.graphpad.com/scientific-software/prism/ | |
Pseudomonas isolation agar | Difco | 292710 | via Fisher Scientific |
Pseudomonas isolation broth | Alpha Biosciences | P16-115 | Custom made batch |
QIAprep Spin Miniprep Kit (250) | Qiagen | 27106 | or preferred method/vendor |
Shaking Incubator | New Brunswick Scientific | Innova 4080 | shake at 200 rpm |
SimpliAmp Thermal Cycler | Applied Biosystems | A24811 | |
Skim Milk | Difco | DF0032-17-3 | via Fisher Scientific |
Small Plates (100 O.D. x 10 mm) | Fisher Scientific | FB0875713 | |
SmartSpec Plus Spectrophotometer | Bio-Rad | 170-2525 | or preferred method/vendor |
Sucrose | Fisher Scientific | S5-500 | |
Toothpicks, round | Diamond | Any brand of toothpicks, autoclaved | |
TOPO TA Cloning Kit, for seqeuncing | Invitrogen | 45-0030 | |
XAF-8 Anesthesia System Filters | Perkins and Elmer | 118999 | |
XGI 8 Gas Anesthesia System | Caliper Life Sciences/Xenogen |