As our project revolves around its potential application in human beings, ensuring patient safety remains a primary focus. Several critical decisions were made with this objective in mind. One of the foremost considerations pertained to the design of the phage delivery system. Specifically, the choice to employ a bacteriophage as the delivery vector was of paramount importance. The bacteriophage possesses a unique characteristic, wherein it can undergo replication solely in the presence of both the phagemid (BBa_K4727003) and the helper plasmid (BBa_K4727002) within the same cell. Notably, the phagemid lacks the genes responsible for encoding the viral capsid, consequently, once the phage particles are produced in the packaging cells, purified, and administered to the patient, they are unable to replicate within the bacterial hosts. Instead, they function solely as carriers to deliver the desired genetic code. This design approach adheres strictly to the most rigorous biosafety guidelines and aligns with bioethical resolutions by preventing the unintended and uncontrolled spread of engineered constructs [1]. Moreover, the inherent specificity of the bacteriophage tropism contributes to precise targeting of specific bacterial strains. Unlike uncharacterized phage cocktails and broad-spectrum antibiotics, this system selectively targets only the intended bacteria, ideally without interfering with the natural microbiota. This specificity adds an additional layer of safety and efficiency to the proposed platform, at least from the point of view of containing recombinant DNA exchange only between vector and target. It is worth underlining that the usage of engineered bacteriophage as it has been conceived in our project, is and adjuvant therapy with antibiotic usage as endpoint; the latter, will in any case harm the whole microbiota, but in this case also including AMR strains.
It is pertinent to emphasize the significance of several systematic reviews that analyzed both clinical and animal trials on phage therapy [2, 3], which have provided valuable insights into the therapeutic approach involving bacteriophages. In general, there are no reported cases of severe adverse effects of bacteriophages on the human health, thus affirming the safety of administering phages to humans through various routes [3]. This outcome is not surprising, given the ubiquitous presence of phages in the environment and their substantial role in the commensal human flora[2]. Consequently, considering the co-evolution of humans with high phage titers over countless millennia, it is unlikely for these viruses to elicit adverse immunological effects [2, 4]. Furthermore even in vulnerable populations and immunocompromised patients, no serious adverse effects have been discovered [4].
The aforementioned interpretation, albeit overly simplistic, posits that the immunogenic attributes and unexpected consequences associated with these formulations are not solely contingent upon the presence of bacteriophages. Rather, they are also contingent upon the methodologies employed for their purification and formulation [3, 4]. Particular attention has been directed towards these preparation methods due to their lack of standardization [3, 4]. For instance, the absence of thorough purification from bacterial toxins (cell wall components, enterotoxins, alpha hemolysin and others) stand as a potential hazard to human health as they can be pro-inflammatory [3, 4]. Furthermore, heightened concern has been directed towards the substantial bacterial lysis that presently ensues upon the administration of phage therapy [4]. In both of these contexts, our developed platform confronts these issues, firstly by facilitating the standardization of the purification procedure. This is achieved by virtue of the phage particles being synthesized within Escherichia coli cells, a well-established and extensively utilized entity in biopharmaceutical production. Furthermore, our proposition involves the adoption of a gene silencing approach targeting antimicrobial resistance-related genes, coupled with the non-replicative nature of the phage particles within the host cell. Consequently, the absence of cellular lysis obviates the release of endotoxins into the systemic milieu.
Even though previously stated, phages have been questioned about impacting the immune system both directly and indirectly [3]. In vitro and in vivo studies verified that bacteriophages impact the innate and adaptive immunity, however their role is still unclear [3]. Clinical studies in healthy adults have highlighted an increase in IgG and IgM levels only if the phage cocktail would be administered via intraperitoneal injection[3], therefore antibody production may depend on the route of administration [3]. Nevertheless no adverse effect has been reported in association with these increases in antibody production but still, studies need to be made to further understand these events and the effects of phage therapy on pregnancy, growth and development.
Ultimately, it is crucial to underscore that due to their specificity, bacteriophages are inherently incapable of infiltrating eukaryotic cells. This inherent specificity mitigates any risks to DNA integrity. Even in the scenario where one of the currently proposed or potential variants of tropism determinants were to demonstrate the capability to infiltrate human cells, we find no conceivable risks. This assertion is supported by the fact that the genetic payload carried by the phage would remain inert within human cells. This is attributed not only to the spatial confinement of putative transcriptional machinery within cell nuclei, but also to the regulatory mechanisms governing gene expression, which are tailored for bacterial hosts and remain unrecognized by human cells.
The choice of use of non-lithic phages has been guided by two important imperatives. The first was to achieve full control over the spread of viral particles inside the human body. As a lithic phage completes its cycle phage particles are released as the cell explodes and dies, this allows phages to spread in the surrounding cellular environment and infect other cells. This phenomenon would lead to an uncontrollable diffusion of the engineered particles. In the second instance we did not want to exercise a high selective pressure over the target bacteria as this would result in the natural selection of variants resistant to the phage infection and lead to a similar effect to the one we are trying to face. If our phage system was to cause bacterial cells death it would result in the selection only for the resistant bacteria and thus their proliferation.
The same argument reported previously can be substantiated also here as the choice of the dead Cas9 follows the idea of exercising a limited selective pressure over the infected bacterial cells. The cut that would be made by the Cas9 enzyme would result in a cellular stress and activation of repair mechanisms that can lead to cellular death. Again in this case this series of events can lead to the selection of phage resistant bacteria that will make our approach non effective. Whereas the use of an interference mechanism mediated by dCas) will result in gene silencing not associated with the activation of repair mechanisms or stress related proteins thus not exercising evolutionary pressure over the bacterial cell.
After engaging with medical professionals and specifically with professor Viola, we were concerned about the immunogenicity of the phage particles, and inherently of their safety. To further understand these aspects, we analyzed with appropriate bioinformatic tools the possible immune responses that could arise from our developed system. This analysis is strictly related to the safety and security chapter of our project but has been deeply analyzed in the modeling section of this wiki.
Briefly we started by researching if there were any known linear epitopes of any M13 deposited protein in the database. We carried out this work utilizing the Immune Epitope Database & Tools (IEDP)[5] and the epitope prediction algorithm associated. The IEDB database had only G3P and G8P proteins registered, thus we could not assess the immunogenicity of the other proteins. The MHC class I complex is present on the surface of almost any human cell, except for immune cells. It binds intracellular pathogen peptides and exposes them to CD8+ T cells. Although MHC-I analysis is not much relevant for our project since the phage does not infect any human cell, we decided to look for MHC-I epitopes anyway because of the higher accuracy of those predictions, compared to the MHC-II epitopes predictions.
The MHC class II complex is present on the surface of antigen presenting cells. It binds internalized pathogen peptides and exposes them to CD4+ T cells. Since the phage tropism spares with high confidence any human cells, MHC-II antigen processing and binding affinity is the focus of our interest.
We then used another software to assess what linear epitopes of the phage proteins are immunogenic for T cells. For this purpose, we used the CD4 T cell immunogenicity prediction software [6, 7], as it allows us to assess the immunogenicity for all seven most common alleles in the population[8], for each query. The software requires the protein sequence as input; the user can predict the T cell immunogenicity using the 7-allele method[8], immunogenicity method and combined method (IEDB recommended). The combined method predicts the final score that combines the predictions from the 7-allele method and immunogenicity method. In order to reach the highest reliability possible, we then compared the results of the two softwares (MHC-Il binding prediction tool and CD4 T cell immunogenicity prediction tool), and thus the predicted epitope through their respective parameter (MHC-II binding prediction and T-cell immunogenicity prediction).Our assumption was that the two software evaluate two different processes of the human immune response. On one hand we have MHC processing, on the other hand we have the binding affinity between epitopes and T cell receptors, thus the grade of probability to activate the T cells. In fact, it was our belief that a comparative analysis between the results of two software could lead to a more reliable prediction.
We used two different softwares to predict B cell epitopes namely DiscoTope and Elliscope. Whereas the former predicts only three dimensional epitopes, with the latter we were able to predict linear epitopes too.
Overall, the results we obtained, given an 80% homology prediction, gave no positive results for allergies and autoimmune diseases. We couldn’t be able to predict, out of the two known epitopes we found, any possible allergic or autoimmune reaction thus suggesting us that our delivery system is to be considered reasonably safe.
Though our system is intended to work in synergy with commonly used antibiotics, we wanted to assess the specificity of the bacteriophage tropism. It is a matter of importance for us that phage infection occurs in a precise manner as we do not want engineered DNA constructs to be unintendedly spread in the environment and in the human body. For this reason, as discussed in the “Specificity assay” chapter in the results section we designed a simple yet informative experiment. After infecting E. coli TOP10 F’, E. coli TOP10 and A. baumannii with our produced phage particles we plated them on selective LB agar plates. The results, clearly shows how the E. coli TOP10 F’ strain has been infected, as expected from the M13 tropism whereas the E. coli TOP10 didn’t grow on the selective plate, meaning that there has not been phage transduction. The analysis of the A. baumannii plate was complex as the strain we tested was intrinsically resistant to ampicillin. Though, there could not be seen any reporter gene expression (namely RFP) so we could conclude that no infection occurred.
We would like to underline that a better and more precise characterization could, and should, be made using more precise techniques as colony PCR to assess the presence of the reporter gene. For matters of time we could not fully complete this part.
After laboratory in-vitro tests, approval from institutions will be required to proceed with evaluations on clinical isolates, to perform toxicity essays on human and/or animal cell lines and pre-clinical tests on animal models. After proving in vitro efficiency and in vivo safety, it would be necessary to develop an adequate formulation of the therapeutic and proceed in testing the developed phage engineered platform on patients affected by AMR pathogens, comparing standard therapies to a conjugated therapy with the engineered phage. The trials are strictly regulated by national and international agencies and are divided in three phases[9]:
Once all these steps are concluded and the data are evaluated the therapeutic would receive the final approval from the regulatory agencies.
Microorganisms are categorized into four biosafety levels based on their potential risk to the environment and human health:
In our project, we chose to work with non-model organisms, specifically Acinetobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa, as they are directly associated with the antimicrobial resistance (AMR) phenomenon. Our objective was to evaluate the efficacy of the CRISPR guides designed for specific targets and to contribute new standard parts and robust protocols to the iGEM community for efficient work with these bacteria. Due to the high risks associated with working with clinical isolates, we opted for safe alternative options. The three bacteria were procured from ATCC, which serves as a standardization organization. The selected strains are considered biosafety level 2 organisms. Specifically, we utilized the following strains:
Given that the adopted microorganisms are designated as "BL-2 laboratory strains", we anticipate no risks or hazards beyond those typically associated with working with biosafety level 2 species. The culture collection data sheets for these strains are available online on the provided product page. Throughout all procedures, we adhered to strict safety measures and utilized appropriate personal protective equipment, including gloves, lab coats, sanitizers, hoods, and other necessary PPE for conducting experiments in a biosafety level 2 laboratory.
Furthermore, our team included clinicians, such as Professor Claudia Del Vecchio, who works in the microbiology and virology unit at Padova University Hospital. The presence of these clinicians ensured continuous access to expert advice and support from institutional biosafety officers, allowing us to address any potential risks related to clinical microbiology implications effectively.
The University of Padova and its laboratories comply with the laws and regulations for biosafety and biosecurity as they are specified in the
All the members of the team involved in wet lab activities have attended a 12 hours safety course for "High risk management" held by the University of Padova where the correct procedures for risk management were clearly explained and before entering in the laboratories all members took a written test assessing their knowledge. The course has been designed and created on the basis of the risks and prevention, protection and management measures of emergencies that generally occur in teaching, research and analysis laboratories. The most relevant chapters for us were: hazardous chemicals (chemical agents and health, health and prevention, acute and chronic toxicity, incompatible agents, REACH and CLP regulations, hazard statements, indication of prudence, correct labeling, safety data sheets, explosive chemical agents, inflammable chemical agents, carcinogenic agents and mutagens) safety criteria in chemical and microbiological laboratories (use of chemical and biological hoods behavior in biological lab, operative safety instructions of machines, equipment, instruments, and correct use of them)