Initial Steps for the Development of a Phage-Mediated Gene Replacement Therapy Using CRISPR-Cas9 Technology

p53 gene (TP53) replacement therapy has shown promising results in cancer gene therapy. However, it has been hampered, mostly because of the gene delivery vector of choice. CRISPR-Cas9 technology (clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9) can knock out the mutated TP53 (mutTP53), but due to its large size, many viral vectors are not suitable or require implemented strategies that lower the therapeutic efficiency. Here, we introduced a bacteriophage or phage-based vector with the ability to target cancer cells and aimed to investigate the feasibility of using this vector to deliver CRISPR-Cas9 transgene in human lung adenocarcinoma cells. First, we produced a tumour-targeted bacteriophage carrying a CRISPR-Cas9 transgene cassette. Next, we investigated any negative impact on vector titers via quantitative polymerase chain reaction (qPCR) and colony-forming agar plate. Last, we combined Western blot analysis and immunofluorescence staining to prove cell transduction in vitro. We showed that the tumour-targeted bacteriophage can package a large-size vector genome, ~10 kb, containing the CRISPR-Cas9 sequence without any negative impact on the active or total number of bacteriophage particles. Then, we detected expression of the Cas9 in human lung adenocarcinoma cells in a targeted and efficient manner. Finally, we proved loss of p53 protein expression when a p53 gRNA was incorporated into the CRISPR-Cas9 phage DNA construct. These proof-of-concept findings support the use of engineered bacteriophage for TP53 replacement therapy in lung cancer.


Introduction
Lung adenocarcinoma is one of the most frequent causes of cancer-related deaths worldwide. Based on data from the World Health Organization (WHO), in 2018 there were approximately 2.09 million cases and 1.76 million deaths. Unfortunately, current treatments have shown few benefits and are the main reason for lung cancer-related deaths due to their aggressive nature.
Gene therapy has emerged as a promising tool to tackle a wide range of diseases, including cancer. Identifying tumour-suppressor genes has been pivotal for the development of anticancer therapies. In particular, the tumour suppressor p53 gene (TP53) has been regarded with interest by many gene therapists since its discovery in 1979 [1]. High throughput screenings in lung cancer patients show that half of tumours present mutations in the TP53 (mutTP53) [2][3][4]. The concept of "TP53 gene replacement therapy" consists of restoring the wild-type (wt) TP53 gene as a means to suppress tumour growth and progression by delivery and expression of a wt TP53 gene in tumours. Reported studies have shown

M13 (Non-Targeted) and RGD4C (Targeted) Bacteriophage-Based Vector Production and Titration
Vectors were generated following a phage-based vector production, as we reported [11,13,20]. The phage capsid contained an additional recombinant r-pVIII major coat protein for the display of H5WYG (GLFHAIAHFIHGGWHGLIHGWYG) histidylated fusogenic peptide with endosomal escape capacity, derived from the N-terminal of the HA2 subunit of the human influenza virus hemagglutinin [13]. Viral particles were produced in Escherichia coli TG1 strain and sterile-filtered through a 0.45μM filter [11]. The total number of particles was measured by traditional qPCR (quantitative Polymerase Chain Reaction). Analysis was performed using the CFX96 Touch Real-Time PCR Detection System. PCR reactions were performed in 15 μL of final volume using the PowerUp™ SYBR™ Green Master Mix (Thermo Fisher Scientific, UK), supplemented with 100μM sense primer targeting AAV2 ITR (5′-GGAACCCCTAGTGATGGAGTT) and 100μM antisense primer targeting AAV2 ITR (5′-CGGCCTCAGTGAGCGA) [25]. A standard curve was prepared using known concentrations of ITR-containing plasmid, and samples were diluted 1:5000 in sterile water. The PCR cycling parameters contained an initial denaturation at 98 °C for 3 min, then 98 °C for 15 s and annealing or extension at 58 °C for 30 s. The plate was read and step 3 was repeated 39 times, followed by a melt curve stage. Data analysis was performed using the Bio-Rad CFX Manager v3.1 and expressed as genome-containing particles/μL. Active particles were titrated based on an agar plate counting system [11,17] and expressed as bacterial transducing units (TU/μL).

Transfection of the Cas9 Phage Plasmid with Polyethyleneimine (PEI)
Cells were grown in Serum Free Media (SFM) for 1 h prior to transfection. A mix of 1:3 ratio (μg DNA: μg PEI (QBiogene, Carlsbad, CA, USA)) was prepared in SFM, vortexed and incubated for 10 min at room temperature (RT). Then the transfection mix was added dropwise into each well of cells. After 3 h post-transfection, the transfection media was replaced with complete DMEM. Map of the CRISPR-Cas9 (clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9) transgene cassette, 7620 bp, inserted within the M13 phage genome. Left and right ITRs (Inverted Terminal Repeats), from AAV2, flank the CRISPR-Cas9 cassette. The guide-RNA (gRNA) scaffold incorporates a target site to make CRISPR-Cas9 protein functional; there is a~1.8 kb space between U6 promoter and the gRNA scaffold. Image generated by SnapGene (GSL Biotech, Chicago, IL, USA).

M13 (Non-Targeted) and RGD4C (Targeted) Bacteriophage-Based Vector Production and Titration
Vectors were generated following a phage-based vector production, as we reported [11,13,20]. The phage capsid contained an additional recombinant r-pVIII major coat protein for the display of H5WYG (GLFHAIAHFIHGGWHGLIHGWYG) histidylated fusogenic peptide with endosomal escape capacity, derived from the N-terminal of the HA2 subunit of the human influenza virus hemagglutinin [13]. Viral particles were produced in Escherichia coli TG1 strain and sterile-filtered through a 0.45µM filter [11]. The total number of particles was measured by traditional qPCR (quantitative Polymerase Chain Reaction). Analysis was performed using the CFX96 Touch Real-Time PCR Detection System. PCR reactions were performed in 15 µL of final volume using the PowerUp™ SYBR™ Green Master Mix (Thermo Fisher Scientific, UK), supplemented with 100µM sense primer targeting AAV2 ITR (5 -GGAACCCCTAGTGATGGAGTT) and 100µM antisense primer targeting AAV2 ITR (5 -CGGCCTCAGTGAGCGA) [25]. A standard curve was prepared using known concentrations of ITR-containing plasmid, and samples were diluted 1:5000 in sterile water. The PCR cycling parameters contained an initial denaturation at 98 • C for 3 min, then 98 • C for 15 s and annealing or extension at 58 • C for 30 s. The plate was read and step 3 was repeated 39 times, followed by a melt curve stage. Data analysis was performed using the Bio-Rad CFX Manager v3.1 and expressed as genome-containing particles/µL. Active particles were titrated based on an agar plate counting system [11,17] and expressed as bacterial transducing units (TU/µL).

Transfection of the Cas9 Phage Plasmid with Polyethyleneimine (PEI)
Cells were grown in Serum Free Media (SFM) for 1 h prior to transfection. A mix of 1:3 ratio (µg DNA: µg PEI (QBiogene, Carlsbad, CA, USA)) was prepared in SFM, vortexed and incubated for 10 min at room temperature (RT). Then the transfection mix was added dropwise into each well of cells. After 3 h post-transfection, the transfection media was replaced with complete DMEM.

Cell Transduction by Phage Vectors
The procedure is based on our previous protocol for cell transduction by phage vectors [17]. Briefly, HEK293 and A549 cells were plated at 5 × 10 5 cells/well in a poly-D-lysine coated 12-well plate. Phage viral stock was diluted in serum-free medium (SFM) to achieve the desired Transducing Unit per cell ratio (TU/cell). The mix was incubated at RT for 15 min. Cells were transduced by drop wising the mix and overnight (O/N) incubation. The next day, the transducing medium was replaced with 500µL complete DMEM. Untreated cells were used as a control. The culture continued for 6 days before carrying out immunocytochemistry.

Western Blot
Cell lysates were prepared using 100 µL 1× SDS sample buffer, followed by sonication for 20 s. Samples were heated at 95 • C for 5 min, cooled on ice and centrifuged for 5 min before loading to SDS-PAGE and transferred to a nitrocellulose membrane. The membrane was blocked with 5% skimmed milk for 1 h. The primary antibodies used were mouse anti-Cas9 (1:1000; Cell Signalling Tech, London, UK) and rabbit anti-GAPDH (1:5000; Santa Cruz Biotechnology, Texas, USA).

Statistical Analysis
To evaluate any potential alteration in viral production due to CRISPR-Cas9 genome vector size, we compared the total and active particle of the former vector and a control. The experiment was repeated 3 times (n = 3) independently, and each replicate was measured 3 times. Statistical analysis was subjected to paired Student's t test with significance at * p < 0.05.

Positive Validation of the CRISPR-Cas9 Phage Plasmid
Plasmid transfection experiments were first performed to validate the design of the CRISPR-Cas9 construct prior to phage vector production and phage cell transduction. Therefore, the A549 human lung adenocarcinoma cells were transfected with the phage plasmid construct and immunocytochemistry was carried out 48 h post-transfection. As shown in Figure 2, cells transfected with the CRISPR-Cas9 phage plasmid expressed the Cas9 protein as compared to control non-transfected cells or cells incubated with a secondary antibody only. To further validate and confirm expression of Cas9 from the phage plasmid construct, we transfected HEK293 cells as a cellular model to maximize protein production and performed a Western blot analysis. The data revealed expression of the Cas9 protein at the expected 160 kDa size ( Figure 3).

Analyzing the Impact of the CRISPR-Cas9 Transgene Cassette on Phage Vector Production and Titers Shows No Significant Alterations
The large size of the CRISPR-Cas9 transgene cassette has been a challenge for vectors to accommodate and deliver [26][27][28]. We therefore sought to carry out quantitative analyses of the produced phage vector particles carrying the CRISPR-Cas9 cassette before cell transduction experiments, in order to examine possible consequences on phage vector production. We evaluated effects on the total vector genome copies and total active phage particles. Our data showed that no changes occurred in the production of active transducing phage and vector genome copies, irrespective of the genome size packaged by the phage capsid ( Figure 4). Indeed, the overall active transducing particles and vector genome of phage carrying CRISPR-Cas9 were similar to that of a phage with a significantly reduced size genome containing a reporter luciferase-based transgene To further validate and confirm expression of Cas9 from the phage plasmid construct, we transfected HEK293 cells as a cellular model to maximize protein production and performed a Western blot analysis. The data revealed expression of the Cas9 protein at the expected 160 kDa size ( Figure 3). To further validate and confirm expression of Cas9 from the phage plasmid construct, we transfected HEK293 cells as a cellular model to maximize protein production and performed a Western blot analysis. The data revealed expression of the Cas9 protein at the expected 160 kDa size ( Figure 3).

Analyzing the Impact of the CRISPR-Cas9 Transgene Cassette on Phage Vector Production and Titers Shows No Significant Alterations
The large size of the CRISPR-Cas9 transgene cassette has been a challenge for vectors to accommodate and deliver [26][27][28]. We therefore sought to carry out quantitative analyses of the produced phage vector particles carrying the CRISPR-Cas9 cassette before cell transduction experiments, in order to examine possible consequences on phage vector production. We evaluated effects on the total vector genome copies and total active phage particles. Our data showed that no changes occurred in the production of active transducing phage and vector genome copies, irrespective of the genome size packaged by the phage capsid ( Figure 4). Indeed, the overall active transducing particles and vector genome of phage carrying CRISPR-Cas9 were similar to that of a phage with a significantly reduced size genome containing a reporter luciferase-based transgene

Analyzing the Impact of the CRISPR-Cas9 Transgene Cassette on Phage Vector Production and Titers Shows No Significant Alterations
The large size of the CRISPR-Cas9 transgene cassette has been a challenge for vectors to accommodate and deliver [26][27][28]. We therefore sought to carry out quantitative analyses of the produced phage vector particles carrying the CRISPR-Cas9 cassette before cell transduction experiments, in order to examine possible consequences on phage vector production. We evaluated effects on the total vector genome copies and total active phage particles. Our data showed that no changes occurred in the production of active transducing phage and vector genome copies, irrespective of the genome size packaged by the phage capsid ( Figure 4). Indeed, the overall active transducing particles and vector genome of phage carrying CRISPR-Cas9 were similar to that of a phage with a significantly reduced size genome containing a reporter luciferase-based transgene cassette (5274 bp) (Figure 4). These findings confirm that neither the phage vector titer nor the fraction of active functional phage were affected by cloning of the CRISPR-Cas9 transgene cassette. cassette (5274 bp) (Figure 4). These findings confirm that neither the phage vector titer nor the fraction of active functional phage were affected by cloning of the CRISPR-Cas9 transgene cassette.

Evaluation of Cas9 Delivery to Lung Cancer Cells by Bacteriophage Vector Carrying the CRISPR-Cas9 Cassette Shows Efficient Transduction
Next, we assessed the potential of RGD4C bacteriophage to target and deliver CRISPR-Cas9 in A549 human lung cancer cells. As shown in Figure 5, we detected Cas9 expression in up to 60-70% of A549 cells at day 6 post treatment with the targeted vector. No expression was detected in cells treated with the non-targeted M13 bacteriophage-based vector carrying the CRISPR-Cas9 transgene cassette. These findings prove that delivery of CRISPR-Cas9 by the bacteriophage vector is efficient and selective, mediated by the RGD4C ligand.

Evaluation of Cas9 Delivery to Lung Cancer Cells by Bacteriophage Vector Carrying the CRISPR-Cas9 Cassette Shows Efficient Transduction
Next, we assessed the potential of RGD4C bacteriophage to target and deliver CRISPR-Cas9 in A549 human lung cancer cells. As shown in Figure 5, we detected Cas9 expression in up to 60-70% of A549 cells at day 6 post treatment with the targeted vector. No expression was detected in cells treated with the non-targeted M13 bacteriophage-based vector carrying the CRISPR-Cas9 transgene cassette. These findings prove that delivery of CRISPR-Cas9 by the bacteriophage vector is efficient and selective, mediated by the RGD4C ligand.
J. Clin. Med. 2020, 9, x FOR PEER REVIEW 6 of 11 cassette (5274 bp) (Figure 4). These findings confirm that neither the phage vector titer nor the fraction of active functional phage were affected by cloning of the CRISPR-Cas9 transgene cassette.

Evaluation of Cas9 Delivery to Lung Cancer Cells by Bacteriophage Vector Carrying the CRISPR-Cas9 Cassette Shows Efficient Transduction
Next, we assessed the potential of RGD4C bacteriophage to target and deliver CRISPR-Cas9 in A549 human lung cancer cells. As shown in Figure 5, we detected Cas9 expression in up to 60-70% of A549 cells at day 6 post treatment with the targeted vector. No expression was detected in cells treated with the non-targeted M13 bacteriophage-based vector carrying the CRISPR-Cas9 transgene cassette. These findings prove that delivery of CRISPR-Cas9 by the bacteriophage vector is efficient and selective, mediated by the RGD4C ligand.

Incorporation of a p53 gRNA in the CRISPR-Cas9 Phage Plasmid Generates Knockout Cells
Our initial experiments showed an accumulation of the Cas9 protein in the cytoplasm, both after vector DNA transfection and phage transduction. However, to exert its function, the CRISPR-Cas9 complex needs to reach the cell nucleus to achieve a targeted gene knockout. Consequently, we cloned a p53 gRNA in the CRISPR-Cas9 phage construct to test its effectiveness in transfected HEK293 cells. First, we carried out immunofluorescence staining using an antibody against p53, and proved expression and nuclear localization of the p53 protein ( Figure 6); which was confirmed by merging with the DAPI stain commonly used as a nuclear counterstain in fluorescence microscopy. Next, importantly, we found that cells displaying nuclear expression of the CRISPR-Cas9, lacked or had negligible p53 protein expression (Figure 7a,b). In contrast, p53 expression was clearly detected in cells lacking expression of the CRISPR-Cas9 (Figure 7b). Figure 5. (a) Confocal images of immunofluorescence staining of Cas9 in A549 cells following transduction with bacteriophage-based vectors at 2,000,000 TU/cell. Cells were treated with either non-targeted (M13) or targeted (RGD4C) vectors carrying the CRISPR-Cas9 cassette. At day 6 post vector treatment, cells were stained with anti-Cas9 (green) and DAPI (blue). Cells incubated with the secondary antibody only were used as control. Scale bar, 154.4μm; (b) Higher magnification image of Cas9 staining in A549 cells upon treatment with RGD4C phage vector carrying the CRISPR-Cas9. Image is separated in channels: Cas9 (green), DAPI (blue) and merged. Scale bar, 77.2μm.

Incorporation of a p53 gRNA in the CRISPR-Cas9 Phage Plasmid Generates Knockout Cells
Our initial experiments showed an accumulation of the Cas9 protein in the cytoplasm, both after vector DNA transfection and phage transduction. However, to exert its function, the CRISPR-Cas9 complex needs to reach the cell nucleus to achieve a targeted gene knockout. Consequently, we cloned a p53 gRNA in the CRISPR-Cas9 phage construct to test its effectiveness in transfected HEK293 cells. First, we carried out immunofluorescence staining using an antibody against p53, and proved expression and nuclear localization of the p53 protein ( Figure 6); which was confirmed by merging with the DAPI stain commonly used as a nuclear counterstain in fluorescence microscopy. Next, importantly, we found that cells displaying nuclear expression of the CRISPR-Cas9, lacked or had negligible p53 protein expression (Figure 7a, b). In contrast, p53 expression was clearly detected in cells lacking expression of the CRISPR-Cas9 (Figure 7b).

Discussion
Here, we utilized a rational design approach to construct and characterize a genetically engineered M13 bacteriophage to deliver in vitro the CRISPR-Cas9 transgene into A549 human lung adenocarcinoma cells. This vector has proved safe in mice and pet dogs upon single and repeated intravenous administrations [11,18,23]. Interestingly, our phage-based vectors are immunogenic but they can be repeatedly administered to efficiently reach their target tissue with no off-target effects

Discussion
Here, we utilized a rational design approach to construct and characterize a genetically engineered M13 bacteriophage to deliver in vitro the CRISPR-Cas9 transgene into A549 human lung adenocarcinoma cells. This vector has proved safe in mice and pet dogs upon single and repeated intravenous administrations [11,18,23]. Interestingly, our phage-based vectors are immunogenic but they can be repeatedly administered to efficiently reach their target tissue with no off-target effects following systemic delivery [11,18,23]. The vector, called RGD4C bacteriophage, can effectively package a 10 kb vector genome. Indeed, no significant reduction was observed in either the active or total number of particles compared to the RGD4C bacteriophage carrying a luciferase transgene cassette with half the size of CRSPR-Cas9 cassette. It is noteworthy to mention that the phage DNA construct already includes a space for the gRNA target sequence, ensuring that the vector genome will not increase in size after cloning the remaining elements. In the gene therapy field, AAV vectors are among the most investigated due to their safety profile. However, in clinical trials for Haemophilia B, patients showed symptoms of inflammation and liver toxicity when large doses of the virus were injected [29]. In addition, the infection efficiency often remains low and unfortunately, due to its immunogenicity, repeated administrations are hindered by neutralizing antibodies [30]. Last, but not least, advanced strategies are required to pack and deliver large size transgenes like CRISPR-Cas9, such as delivering the gRNA by a different vector, which reduces the genome size but also compromises the efficiency of therapy.
The original aim of the study was to introduce RGD4C bacteriophage as a potential gene delivery vector in lung cancer gene editing. Altogether, our results indicate that the vector can successfully deliver the CRISPR-Cas9 transgene cassette and express the Cas9 protein in human lung adenocarcinoma cells efficiently and selectively. We also showed that the transgene cassette is functional and able to knock out p53 protein expression after adding the appropriate gRNA.
Ideally, following these initial steps, we would then include in the same vector a healthy copy of TP53 to correct the aberrant gene. TP53 has a major role in deciding the fate of a cell, such as cell death or cell cycle arrest. Somatic mutations in TP53 can be decisive for the commencement of tumorigenesis, progression or malignancy. Hence, it is no doubt that p53 is a hot topic in cancer gene therapy. Challenges lie ahead on the complexity of the therapy. The basis of TP53 gene replacement therapy consists of gene editing the mutTP53 allele and replacing it with a healthy one. This requires gene editing tools like the CRISPR-Cas9 technology to cut out the aberrant allele and a wt TP53 gene to replace it. Hence, viral vectors with small gene packaging size are not suitable.
Our findings, together with other features, suggest that RGD4C bacteriophage is a promising model for TP53 gene replacement therapy. A change in the promoter choice and structural modifications of the phage capsid are already taken into consideration for an improved generation of the vector [13,23].
This project may have an interesting impact in the clinical development of new gene therapies. The RGD4C bacteriophage has shown safe systemic administration, tolerability to repeated administrations, tumour tropism and cost-effectiveness [18,19,22,[31][32][33][34]. These characteristics, in addition to its high packaging size, bring a new competitive gene delivery vector to the field of CRISPR-Cas9 technology.

Conclusions
Our RGD4C bacteriophage-based vector can carry genome sizes superior to 10 kb without any significant reduction in its titers. This includes the CRISPR-Cas9 phage vector (10,183 bp) constructed in this study for a future lung cancer TP53 gene replacement therapy. The vector can target A549 human lung adenocarcinoma cells in vitro and deliver the CRISPR-Cas9 sequence selectively with a 60-70% efficiency.