T4 phage do not activate the intracellular DNA-sensing receptors TLR9 and cGAS-STING

We focused on bacteriophage T4, a virulent Tevenvirinae phage, with an approximately 200-nm long myovirus morphology and a 168,903-bp genome, which infects Escherichia coli [33,34]. This phage was selected as it was previously demonstrated to be internalized by mammalian cells, accumulating intracellularly within macropinosomes over time [12,13]. Phage lysates were purified and concentrated via ultrafiltration following the Phage on Tap protocol to produce a single, high-titer phage stock that was used for all subsequent assays [31,32]. Phage stocks were treated with DNase and RNase to remove extracellular nucleic acids followed by endotoxin removal using 1-Octanol washes. As bacterial endotoxins are known to trigger an innate immune response in TLR4 expressing mammalian cells, we ensured all phage samples were deplete of endotoxin (<1 EU/mL). Despite this, there remained the possibility of bacterial components (i.e., proteins, polysaccharides, nucleic acids) persisting at low levels within our phage lysates. To address this, we passed the phage lysate through a 0.02-μm filter 4 times to generate a phage-free lysate that would also contain any residual bacterial components, henceforth referred to as “Filter control,” which served as a comparative control to ensure cellular responses were phage driven and not induced by any bacterial residues or buffer contaminants. We further prepared 2 additional samples, referred to as “Capsid-only” and “Phage DNA.” The Capsid-only sample was prepared by successive heat treatment of T4 phages to break the capsid followed by DNase treatment to eliminate the DNA and thus contains a mixture of partially degraded phage proteins. The Phage DNA sample was prepared via DNA extraction using a column to produce a phage-genome sized band when visualized on an agarose gel, with further DNA integrity checked by T4-specific PCR.

We then investigated whether T4 phages and associated controls could be internalized by our in vitro tissue culture cells, and whether they activate key intracellular nucleic acid receptors, which stimulate downstream pro-inflammatory immune pathways. T4 phages were applied to both A549 human lung epithelial cells and MDCK-I dog kidney epithelial cell lines and were visualized as distinct fluorescent puncta within the cell cytoplasm, suggestive of phage accumulation within membrane-bound vesicles (Fig 1A and 1B) [12,13]. Once internalized, T4 phage and their genomic material could be recognized by the nucleic acid receptor TLR9 [35], a transmembrane protein that resides within endocytic vesicles and preferentially binds DNA from bacteria and viruses [21]. Once activated, TLR9 leads to a downstream cascade via the MyD88 pathway, resulting in the induction of inflammatory cytokines through activation of NF-κB and other transcription factors, including IRF7, which bind the IFN-β, promoter [36–39]. To test this, we used a luciferase-based luminescence assay to detect the downstream activation of TLR9 in activating the IFN-β promoter in A549 cells following the addition of either T4 phage or the Filter control [40]. We saw no activation of luciferase expression in either the T4 phage or Filter control samples, while transfection of the positive controls showed strong activation of expression from both plasmids (Figs 1C, 1D and S1). From these results, we conclude that neither NF-κB-dependent activation, nor activation by other elements that activate IFN-ß expression were induced by the internalization of T4 phages. This suggests that the T4 phage capsid remains intact and that phage DNA was not exposed nor detected by TLR9 within the macropinosome.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Uptake of T4 phage by mammalian cells does not trigger a pro-inflammatory immune response.

(A) A549 cells and (B) MDCK-I cells incubated with T4 phages for 2 hours. Images were taken with a confocal microscope; the plasma membrane is shown in magenta, T4 phage DNA in green, and the cell nucleus in blue. (C) A549 cells transfected with NF-κB-dependent luciferase reporter plasmid, or (D) IFN-β promoter-dependent luciferase reporter plasmid, followed by 48 hours incubation with 10⁹ T4 phages/mL or a Filter control. Differentiated WT (E) or STING KO (F) BMDM cells were incubated for 18 hours 10⁷ T4 phages/mL, Filter control, Capsid-only or transfected with phage DNA using Lipofectamine 2000. Raw data can be found in S1 Data; each set of data follows the normality law; P values between the different groups were calculated from a one-way ANOVA with multiple comparisons, shown as stars (P < 0.0001 = ****; A: F (3, 12) = 31.06; B: F (3, 12) = 2.812; C: F (4, 25) = 5.7; D: F (4, 25) = 0.8181).

https://doi.org/10.1371/journal.pbio.3002341.g001

Following internalization and trafficking, phage particles or DNA may escape the macropinosome and gain access to the cytosol. Here, the presence of unencapsulated phage DNA would be recognized by the cGAS-STING pathway, leading to the production of IFN-β and inflammatory cytokines [41–44]. To test this, we incubated wild-type (WT) bone marrow-derived macrophages (BMDMs) or STING knock-out (KO) BMDM with either 10⁷ T4 phages/mL, Filter control, or Capsid-only samples for 18 hours. After incubation, the activation of the cGAS-STING pathway was measured by ELISA to measure IFN-β levels (Fig 1E and 1F). We saw no STING induction in either the Phages, Filter control, or Capsid-only samples. As an additional a positive control, we transfected cells with extracted T4 phage DNA using Lipofectamine 2000 to demonstrate that phage DNA can activate cGAS-STING, with phage DNA showing a strong activation of STING in the WT cells (Fig 1E). Comparatively, in the STING-KO BMDM cells, we did not see any activation of the cGAS-STING pathway for any of the controls or samples (Fig 1F). Importantly, both WT and STING-KO BMDM cells also have a functional TLR9, suggesting that the fact our STING KO cell lines did not respond to phage stimulation provides further confirmation that TLR9 is indeed not in play. In summary, highly purified T4 phage were internalized by mammalian cells but did not activate pathways downstream of TLR9, including NF-κB-dependent pathways, nor cGAS-STING signaling pathways. This suggests that internalized T4 phages capsids remain intact or are trafficked in such a way as to prevent phage DNA from being exposed and triggering the innate immune system.

T4 phage induces protein expression and phosphorylation changes in cell signaling pathways

We utilized an antibody microarray to investigate broader cellular changes in the expression and phosphorylation of key cell signaling proteins in response to phage T4 compared with the Filter control. We used 2 protein microarrays from Kinexus Biotech, the KAM-1325 microarray that contains 1,325 pan- and phosphosite-specific antibodies covering all the main cellular signaling pathways, and the KAM-2000 microarray with 2,000 pan- and phosphosite-specific antibodies. Importantly, the MDCK-I samples were analyzed by the KAM-1325 antibody array, while the A549 samples were analyzed using the improved KAM-2000 antibody array, which includes most of the antibodies from the prior array, along with 675 additional antibodies for improved detection of cellular changes.

For deconvolution of the microarray datasets, we used the previously developed methodology by Adderley and colleagues [45]. Datasets were mapped onto a network followed by a pathway analysis, which utilizes random walks to identify chains of phosphorylation events occurring more or less frequently than expected [45]. Rather than focusing solely on the largest fold changes, this analysis identifies cellular pathway interactions to provide an interpretation of the most important pathways that are influenced by exposure to phages This provided us with an overview of the main cellular pathways that were influenced by the presence of phages (Fig 2A: MDCK-I and Fig 2B: A549; S2 and S3 Figs). From this analysis, we found 52 hits for the MDCK-I cells, and 150 hits for the A549 cells, which utilized the improved KAM-2000 antibody array. Based on this analysis, we focused our attention on 2 pathways—AKT and CDK1. These pathways were selected as they were common across the 2 antibody microarray datasets and were both associated with long network associations.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Network analysis of mammalian cells treated with T4 phages.

(A) Kinexus KAM-1325 antibody microarray with MDCK-I cells after 8 hours of incubation with T4 phages. (B) Kinexus KAM-2000 antibody microarray with A549 cells after 8 hours of incubation with T4 phages. Figures report major cellular pathways of the main up- and down-regulated leads from the network analysis. Boxes highlighted in red are proteins discussed in this manuscript. The color gradient and arrow width indicate the Log₂ fold change values.

https://doi.org/10.1371/journal.pbio.3002341.g002

T4 phage activates the AKT pathway promoting cell growth, survival, and macropinocytosis

The AKT signaling pathway regulates a myriad of cellular functions, including promoting cell growth, proliferation, survival, and metabolism [46]. AKT itself is a serine/threonine-specific protein kinase that is activated through extracellular growth factors, such as insulin, which are detected through receptor tyrosine kinases (RTKs) or G protein-coupled receptors (GPCRs). These receptors recruit PI3K (also called PIK3CA) to the membrane, leading to the subsequent recruitment of PDK1 (also called PDPK1), which, once activated, will phosphorylate AKT on the tyrosine amino acid, T308. Alternatively, PDK1 may recruit mTORC2, which itself will activate AKT through the S473 phosphorylation site [46,47]. Once activated, AKT and its downstream effectors will induce a broad range of cellular responses, including activation of glycolysis, protein synthesis, cell survival and proliferation, glycogen synthesis, fatty acid synthesis, and the inhibition of autophagy.

In our MDCK-I datasets, we observed the activation of AKT by both PDK1 through the S473, and SRC through the Y326 phosphorylation sites in the presence of T4 phage (Fig 2A). Once activated, AKT led to increased inhibition of BAD (also called BCL2) through the S75 phosphorylation site, which is an agonist of cell death that prevents apoptosis and enhances cell survival [47–49]. Interestingly, in our A549 dataset (Figs 2B and S3), we observed similar activation of AKT through the S473 phosphorylation site indirectly by PDK1, but also via MAPK14, known as p38α MAPK. P38α MAPK is activated through environmental stressors and pro-inflammatory cytokines and usually results in increased cell survival [50]. Further, within the A549 cells, the activation of AKT induced the phosphorylation of EZR (Ezrin) through the T567 phosphorylation site [51,52]. Importantly, EZR acts as an intermediary between the plasma membrane and the actin cytoskeleton of the cell, with its activation being required for the formation of membrane ruffles during macropinocytosis, fusion of cell-to-cell membranes, and the formation of endosomes [53].

T4 phage inhibits the CDK1 pathway to delay cell cycle progression and prolong cellular growth

The activity of cyclin-dependent kinases (CDKs) controls all aspects of cell division. CDK1 is implicated in many, if not all, cell cycle regulatory pathways and is the central hub for regulating cells progressing through the G2 growth proliferation and mitosis phases of the cell cycle [54–56]. CDK1 is essential and sufficient to drive the mammalian cell cycle, including the entry and exit of mitosis and signaling the start of the growth proliferation phase [55,57,58].

In our MDCK-I samples (Figs 2A and S2), we observed an indirect down-regulation of CDK1 at the T161 and Y15 phosphorylation sites. The down-regulation of CDK1 inhibits cells from progressing through the G2 and mitotic phases of the cell cycle. This resulted in cascading down-regulation of other cell cycle effectors, including JUN, which is a transcription factor implicated in the prevention of apoptosis and is responsible for the progression of the cell cycle through the G1 growth phase, via down-regulation at the phosphorylation site S73 [59]. With the reduction in activation of JUN, the progression through the G1 phase of the cell cycle would be delayed, keeping cells in a prolonged state of cellular growth. Simultaneously, we observed the down-regulation in the activation of ribosomal protein p70S6K (S6 kinase beta-1 also called RPS6KB1), which regulates both cell death and proliferation [60]. This down-regulation was mediated through 2 distinct phosphorylation sites, being the T252 phospho-site, which was acted upon by PI3K and PDK1, and S447, which was acted upon by CDK1. Once inhibited, the RPS6KB1-mediated inhibition of IRS-1 (insulin receptor substrate 1) was removed. Previous reports suggest that IRS-1 can further activate PI3K, thereby leading to a positive feedback loop where IRS-1 activates PI3K to further increase AKT activation again [46]. Interestingly, we also observed the down-regulated activation of ESR1 (estrogen receptor α) in both MDCK-I and A549 arrays from the upstream effectors CDK1 and SRC [61],

In the A549 sample (Figs 2B and S3), we observed similar down-regulation of CDK1 as in the MDCK-I array, but here through the in-direct phosphorylation of Y15 site by both SRC and CHEK1 (checkpoint kinase 1). CHEK1 plays an essential role in cell cycle regulation and DNA damage response [62]. CHEK1 further regulates the G1/S transition (along with other cell cycle checkpoints) and is responsible for preventing cells with DNA damage from progressing through the cell cycle [63]. At the same time, we saw that CHEK1 was down-regulating the phosphorylation of the tumor suppressor protein TP53 through the S37 phospho-site [64,65].

Treatment with phage leads to prolonged cell cycle growth phase and increased cellular proliferation

From the microarrays, we observed a common pattern where T4 phages induced cell cycle arrest at the G1 phase and increased cellular metabolism and growth. To validate these responses, we applied T4 phage to A549 cells to investigate differences in cell cycle and proliferation. First, we utilized a comprehensive FACS assay that measured the DNA concentration within each cell to assign them a cell cycle phase, with A549 cells being treated with either T4 phages or the Filter control for 8 or 24 hours before FACS analysis (Fig 3A) [66]. We saw no significant differences in cell cycle between the T4 phage treated and Filter control cells at the 8-hour time point. However, we did observe a significant increase in the proportion of T4 phage-treated cells in the G0/G1 phase of the cell cycle compared with the Filter control at the 24-hour time point (P = 0.0463). This suggests that, in line with our microarray observations, T4 phage application to A549 cells leads to a prolonged G0/G1 phase that would facilitate broad changes in metabolism, cellular growth, and cell survival.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Phage application to in vitro mammalian cells leads to enhanced growth and proliferation.

(A) Cell cycle stage repartition within the A549 cell population after 8- or 24-hour incubation with phages or Filter control (data are mean with error bars representing 95% CI, n = 3 independent replicates with 100,000 cells analyzed). Not all the cells are included in a cell cycle phase (S4 and S5 Figs). P values of each cell cycle stage between the Filter control and T4 phage were calculated using a two-way ANOVA, shown as stars (F (3, 32) = 2.237). (B) Cell proliferation assay as measured via absorbance (540 nm) using a modified MTT colorimetric assay with A549 cells incubated with phages for 24, 48, or 72 hours (data are mean with error bars representing 95% CI, n = 3 independent replicates). Raw data can be found in S2 Data; P values were calculated using a two-way ANOVA, shown as stars (F (2, 224) = 1,015).

https://doi.org/10.1371/journal.pbio.3002341.g003

Next, we performed an MTT cellular proliferation assay, which uses a colorimetric reduction of tetrazolium salt to measure metabolically active cells, as an indicator of cellular proliferation. A549 cells were seeded at a low density (2 × 10⁴ cells/mL) followed by treatment with either T4 phages, the Filter control, or a cell only control (Fig 3B). We observed a significant increase in cell proliferation for the T4 phage-treated samples at 72 hours compared with cell only control (P = 0.032) and the Filter control (P = 0.0044). This suggests that T4 phage application to mammalian in vitro cells leads to enhanced cellular proliferation over a period of days. To extend these findings, we tested an additional 3 phages for their capacity to enhance cellular growth. These include øNS09, an uncharacterized virulent podovirus with a 44-kb genome, phage T3, a virulent podovirus with a 38-kb genome, and Lambda vir, a virulent siphovirus with a 48-kb genome, all of which infect the bacterial host E. coli. While all phages showed a trend for increased cell proliferation compared to the cell only control, only øNS09 was significantly different (P = 0.0289) (Fig 3B). These results suggest that addition of diverse phage to mammalian cells in vitro leads to cascade of cellular responses that promote cellular growth and proliferation.

Bacterial stocks and phage stocks

The bacterial strain used in this study was E. coli B strain HER 1024, which was cultured in lysogeny broth (LB) media (10 g tryptone, 5 g yeast extract, 10 g NaCl, in 1 L of distilled water [dH₂O]) at 37°C shaking overnight and used to propagate and titer T4 phages supplemented with 10 mM CaCl₂ and MgSO₄. T4 phages were cleaned and purified using the Phage on Tap (PoT) protocol [31] and titered up to a concentration of approximately 10¹¹ phages/mL to produce a phage stock solution that was used for all experiments. After purification, phages were treated with DNase-I and RNase and then stored in a final solution of SM Buffer (2.0 g MgSO₄·7H₂O, 5.8 g NaCl, 50 mL of 1 M Tris–HCl (pH 7.4), dissolved in 1 L of dH₂O) at 4°C.

Endotoxin removal

The endotoxin removal protocol followed the PoT protocol [31]. The phages lysate was cleaned 4 times with 1-Octanol to remove endotoxins from the lysate, reducing endotoxins from 5,734 EU/mL to 167 EU/mL in the final phage stock solution (10¹¹ phages/mL) (see also [32]). In all experiments, unless otherwise stated, phages were diluted in endotoxin-free buffers to a final concentration of 10⁸ PFU/mL (unless otherwise stated), resulting in an endotoxin concentration below 1 EU/mL.

Control sample preparation

We prepared several control samples all using the same ultra-pure T4 phage lysate. First, in the Filter control sample, the lysate was passed 4 times through a 0.02-μm filter to remove phage particles from the lysate. The absence of phages was confirmed using a top agar assay at neat dilution with no plaques observed. Second, the capsid-only sample was obtained by breaking the phage capsid using heat treatment. Phages were heated at around 70°C to break open the capsid and release DNA. The sample was then treated with DNase-I to degrade the phage DNA and only the empty capsids remained. Again, the absence of active phages was tested using plaque assay. Finally, the phage DNA sample was obtained using the Norgen Phage DNA isolation kit (Norgen Cat#46800) following the manufacturer’s instructions and confirmed using T4 phage-specific PCR.

Cell line stocks

The choice of A549 and MDCK-I cells was made based upon observations in our previous study demonstrating the internalization of T4 phages across both these cell lines [13]. A549 cells were grown in Ham’s F-12K (Kaighn’s) (also called F12-K) (Life Technologies Australia) media with 10% fetal bovine serum (FBS) (Life Technologies Australia) at 37°C and 5% CO₂ and supplemented with 1% penicillin–streptomycin (Life Technologies Australia). MDCK-I cells were grown in Modified Eagle Medium (MEM) (Life Technologies Australia) with 10% FBS supplemented with 1% penicillin–streptomycin (Sigma-Aldrich, Australia) at 37°C and 5% CO₂.

Confocal microscopy

For the confocal microscopy experiment, cells were seeded in an IBIDI μ-Slide 8-well glass-bottom slide (DKSH Australia). When cells reached 80% to 90% confluency, cells were incubated for 20 minutes with the respective culture media for each cell line with 5% Hoechst 33342 stain, excitation/emission 115,361/497 nm (Life Technologies Australia) and 1% CellMask deep red plasma membrane stain, excitation/emission 649/666 nm (Life Technologies Australia). After incubation cells were washed 3 times with Dulbecco’s phosphate-buffered saline (DPBS) and then left in Hank’s Balanced Salt Solution (HBSS) with 1% FBS until acquisition. Purified phages were labeled with 1% SYBR-Gold, excitation/emission 495/537 nm (Life Technologies Australia), following the protocol in Bichet and colleagues [32]. Around 200 μL of clean stained phages were in each well containing cells, right before the start of the acquisition (see the detailed protocol in Bichet and colleagues [32]). Cells were imaged with HC PL APO 63×/1.40 Oil CS2 oil immersion objective by Leica SP8 confocal microscope on inverted stand with a hybrid detector (HyD) in real time. HyD detector was used in sequential mode to detect the phages. One image was acquired every 2 minutes for 2 hours. Time lapses were created through postprocessing using the FIJI software version 2.0.0-rc-68/1.52f. (See the detailed protocol in Bichet and colleagues [13].)

Luciferase assay

A549 cells were plated at 1.5 × 10⁵ cells/mL in 24-well cell culture plates (Corning) for 1 day. Once cells reached 70% confluency, the cells were cotransfected using Fugene HD transfecting reagent at a 1:3 ratio (Promega) along with pRL-TK Renilla (Renilla Luciferase, internal control; Promega) as an internal transfection control. We used the reporter plasmid pIFN-β-GL3Luc, which carriers the promoter region of IFN-β gene, and pNF-κB-Luc, which contains 5 copies of the NF-κB binding motif of the IFN-β promoter, upstream of a luciferase report gene (Firefly Luciferase) [40]. As positive controls to activate expression from the plasmids, we used FLAG-MAVS and pEF-FLAG-RIGI-I(N), which are known to activate the IFN-β promoter, through activation of NF-κB and parallel pathways, or the negative control pUC-18 (empty-vector). A transfection control well was transfected with peGFPC1 instead of pUC-18 to measure the transfection rate with a fluorescent microscope. Each well was transfected in duplicate. One day posttransfection of the reporter plasmid carrying the luciferase cassette, either T4 phage at a titer of 10⁹ PFU/mL or the Filter control were added to the transfected cells and incubated for 2 days. Cells were then incubated for 30 minutes at 4°C slowing rotating with Passive Lysis Buffer (Promega Cat#E1941). After incubation with the lysis buffers, cells were scraped and collected and spun down for 3 minutes at high speed. The supernatant was collected and kept at −20°C until analysis. The values for firefly luciferase activity were normalized to those of Renilla luciferase by calculating the ration of firefly to Renilla luminescence.

BMDMs isolation and differentiation

BMDMs were obtained by differentiating isolated bone marrow cells from the femurs of the STING-deficient and matched WT control [85]. Briefly, bone marrow cells were flushed, washed, and differentiated in a 20% L929 cell-conditioned medium for 6 days at 37°C in a 5% CO₂, as described previously [86]. The use of mouse tissues was approved by the Monash University Animal Ethics Committee (MARP/2018/067).

cGAS-STING phage assay

Approximately 24 hours before the phage treatment BDMD cells were detached by gently scraping the flasks and plated at 100,000 cells per well in a final volume of 200 μL in a 96-well plate. Around 24 hours after plating the cells, 2 μL of 10⁷ phages/mL solution, Filter control, or capsid-only samples (or 1.3 ng of phages DNA complexed with Lipofectamine 2000 as control) were added to the BMDM cells for another 18 hours. Murine IP-10 production was measured from 100 μL of the supernatant from the BMDMs using Mouse CXCL10/IP-10/CRG-2 Duo Set ELISA (R&D Systems, #Dy466) according to the manufacturer’s protocol.

KAM-1325 and KAM-2000 Kinexus antibody microarrays

Following the protocol described by Bichet and colleagues [32], confluent MDCK-I or A549 cell lines, for KAM-1325 or KAM-2000 microarrays, respectively, were incubated with T4 phages or Filter control samples for 8 hours at 37°C and 5% CO₂. After incubation, the cells were scraped in lysis buffer before sonication. All samples were treated as chemically lysed proteins and followed the recommended protocol by Kinexus. The MDCK-I and A549 proteins were quantified using the Bradford protein concentration assay (Thermo Fisher Scientific). The MDCK-I samples were then incubated on the KAM-1325 array before sending the array to Kinexus for analysis, while the A549 samples were sent directly to Kinexus for analysis on the KAM-2000 array. Briefly, microarray datasets were filtered to remove low signal intensity and/or relatively high error signals compared to control signals. The network analysis was then separately run for both up- and down-regulated phosphorylation events before being assembled into a comparative pathway map [45]. Only pathways with more than 2 intermediates and with fold-changes greater than 5% CFC (% changes from control) were selected for further consideration.

Microarray analysis

The analysis of the microarray dataset was performed using the MAPPINGS V1.0 network analysis program developed by Adderley and colleagues [45]. There was redundancy between the antibodies tested across the KAM-1325 and KAM-2000 arrays with different antibodies targeting the same protein for more precision (S2 and S3 Figs and S1 and S2 Tables). First, the signals were filtered and all signals below 1,000 units were considered as low for the KAM-1325 array and below 500 units for the KAM-2000 array and removed from the assay. Any signal with a high error relative to signal change was disregarded, and any antibody with a higher total signal error across all the arrays compared to the control arrays was disregarded. Next, each unknown substrate effect, where no known biological data were found, was considered as an activation effect for this analysis. All nodes (kinases) without a directed edge toward them were the most probable kinase for the downstream phosphorylation event. This was a consequence of the microarray data not reporting which kinase is responsible for each phosphorylation event. Independent positive and negative network analyses were analyzed, and in the case of parallel phosphorylation, only the ones with the greater magnitude values were selected and appeared on the pathway map. To ensure the proper termination of each pathway, 3 options were chosen: (1) If no other path were available after the last kinase or substrate, then the pathway was stopped. (2) If the last phosphorylation had an inhibitory effect, then the pathway was stopped as well. (3) Finally, for pathways with only one downstream option available but with no changes from the datasets between the cells incubated with the phages and the Filter control, then a percentage of fold change was assigned for each of these single paths, either between 0% and 20% for the experimental dataset and 20% for the control [45].

FACS assay

A549 cells were plated at 8 × 10⁴ cells/mL. Approximately 24 hours after plating the cells, phages at 10⁹ PFU/mL or Filter control samples were added to the cells and incubated for 8 or 24 hours. Cells were then washed twice with PBS; the washes were collected in a 15-mL falcon tube to prevent the loss of dead floating cells and a bias in the analysis. We added trypsin to the wells to collect the cells in the corresponding 15-mL falcon tubes. Cells were quickly centrifuged 3 minutes at 1,500g before adding 1 mL of cold PBS to the pellet. While vortexing, we slowly added 2.3 mL of ice-cold 100% EtOH. Cells were incubated for 40 minutes at 4°C. Cells were centrifuged for 5 minutes at 300g and resuspended in 500 μL of cold PBS. Cells were again centrifuged for 5 minutes at 300g and resuspended in cell cycle buffer (in PBS, add 30 μg/mL of PI (propidium iodide), 100 μg/mL of RNase A), incubating the cells at RT in the dark for 45 minutes. The cells were centrifuged for 5 minutes at 300g and resuspended in 500 μL of cold PBS before being transferred to a 5-mL round polystyrene bottom tube (Corning). Cells were left at 4°C until FACS analysis following the [66] protocol. The experiment was performed with triplicate wells for each condition and a control well with no PI (S4 and S5 Figs). Data were generated on a 4-laser Fortessa X-20, manufactured by Becton Dickinson (BD). A total of 100,000 cells were analyzed for each assay and assigned to a cell cycle stage: G0/G1, G2, S, or Sub G0.

Tetrazolium MTT proliferation assay

For the proliferation assay, a modified MTT (3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide, Thermo Fisher Scientific) colorimetric assay was used [87,88]. A549 cells were plated at 2 × 10⁴ cells/mL in standard 96-well plates (Thermo Fisher Scientific) and incubated at 37°C and 5% CO₂ overnight to recover from handling and allow attachment. The next day, 10 μL of purified phage solution at 10¹⁰ PFU/mL were added to the cells (final concentration 10⁹ PFU/mL), and plates were incubated for up to 72 hours. To account for any influence on cell proliferation potentially due to the phage lysates background, a Filter control, and a cell only control was included. At 24-, 48-, and 72-hour time points, 10 μL of 5 mg/mL MTT solution were added to all wells, including controls (final concentration 0.5 mg/mL). Cells were then incubated at 37°C and 5% CO₂ for 3 hours to permit intracellular formation of purple formazan crystals. Briefly, the formation of formazan is a result of tetrazolium reduction mediated by metabolically active cells and the amount that is formed broadly reflects the total number of proliferating cells. After incubation, cell culture media were removed and 50 μL of pure dimethylsulfoxide (DMSO; Merck) were added to each well to dissolve formazan crystals. To ensure complete dissolution, DMSO was mixed with cells by pipetting thoroughly, taking care to prevent bubble formation. Plates were incubated on an orbital shaker (200 rpm) in the dark for 10 minutes at 37°C. Absorbance values at 540 nm were then determined for each well using a microplate reader (Epoch2, BioTek). The assay was repeated with a total of 3 independent biological replicates with each sample tested in triplicate.

Statistics

n represents the number of samples analyzed; each sample was performed at minimum in duplicate. The datasets were tested for normality before statistical analysis. All the statistics across this article were done using the GraphPad Prism software using one-way ANOVA or two-way ANOVA including Dunnett’s multiple comparison tests. The results of the statistics are represented with stars on top of the corresponding data. All the microarray experiments were performed with only one sample; no statistics were applicable to these results.