Potential of bacteriophage therapy in managing Staphylococcus aureus infections during chemotherapy for lung cancer patients

Respiratory Staphylococcus aureus infection represents a common complication in lung cancer patients, which is characterized with progressively and recurrently intratumor invasion. Although bacteriophages are widely reported as an effective bioweapon for managing bacterial infections, its applicability in handling infectious complications during cancer chemotherapy remains unknown. In this work, we hypothesized cancer chemotherapeutics would influence the efficacy of bacteriophages. To verify this end, interactions between four anticancer drugs (Gemcitabine, Doxorubicin, Cisplatin, and Irinotecan) with phage K were investigated, where Cisplatin directly reduced phage titers while Gemcitabine and Doxorubicin partially inhibited its propagation. The antibacterial efficacy of drug-phage K combinations was tested in a S. aureus infected cancer cell model. Doxorubicin enhanced the antibacterial capacity of phage K, destroying 22-folds of cell-associated bacteria than that of phage K alone use. Also, S. aureus migration was remarkably reduced by Doxorubicin. Overall, our data suggested that Doxorubicin had synergistic effects with phage K in combating S. aureus intracellular infection and migration. This work may broaden the options of indication for phage clinical transformation and also provide reference for the adjunctive application of chemo drugs in intracellular infection management.

www.nature.com/scientificreports/ instrumental to its lytic cycle. Chemo drugs were chosen based on clinical use and characterized by their unique action mechanisms, where Gemcitabine prevents DNA chain elongation 17 , Doxorubicin inhibits the function of topoisomerase II enzyme 18 , Cisplatin breaks the single-stranded DNA 19 , and Irinotecan inhibits topoisomerase I once activated 20 . A549 cancer cell was used to establish a S. aureus invasive cancer cell model. The behavior of phage K in managing S. aureus intracellular invasion and migration was studied, and the impacts of chemo drugs on its antibacterial efficacy were also explored. We meaningfully discovered that Doxorubicin promoted the efficiency in eradicating cell-associated bacteria, which can be further considered as an adjuvant of phage therapy in the treatment of bacterial infections for cancer patients.

Results
Antibacterial activity of phage K/anticancer drugs against S. aureus. Here, the bacteriolytic activity of phage K against S. aureus was first evaluated. As shown in Fig. 1A, the lysis kinetics of phage K to S. aureus was in a combination of MOI-and time-dependent manner. Phage treatments with MOI ranging from 10 -1 to 10 -4 showed considerable reduction in the OD 600 value compared with the control group after 24 h, where MOI 10 -1 displayed the most rapid killing effect. MOI at 10 -5 and 10 -6 , however, exhibited comparable OD 600 values with that of the control group. Fluorescent images (Fig. 1B) also revealed that S. aureus (with an inoculum size of 10 8 CFU/mL) was remarkably destroyed by phage K (MOI at 10 -1 ) after 24 h of treatment. At the end of experiment, no bacterial colony was found in the treatment group, while the bacterial density increased to 10 9 CFU/ mL (marked as green fluorescence) in the control group. Meanwhile, the MICs and antibacterial efficiency of four clinically prescribed anticancer drugs (Gemcitabine, Doxorubicin, Cisplatin and Irinotecan) against S. aureus were determined. Results showed that Doxorubicin had the smallest MIC (5.5 μM), indicating the strongest capability of inhibiting bacterial growth, followed by Gemcitabine with a MIC of 10.0 μM, whereas Cisplatin and Irinotecan exhibited no bacteriostatic effects ( Table 1). The antibacterial effects of these anticancer drugs at the assigned concentration (10 μM), chosen based on the drug cytotoxicity to the cancer cells ( Fig. S1) to ensure sufficient cancer cell survival, were illustrated in Fig. 2A. . Each symbol represents the mean ± SD of three replications. (B) Fluorescent images of S. aureus (10 8 CFU/mL) with PBS (control) and phage K (10 7 PFU/mL) treatments (MOI at 10 -1 ) for 24 h (bacteria grew to 10 9 CFU/mL in the control, and no bacterial colony was found in the treatment). Bacteria was stained with a green fluorescent dye (SYTO™ 9). Bars represent 100 μm. Interactions between phage K, anticancer drugs, and S. aureus. To better weigh the feasibility of applying phage therapy during chemotherapy, interactions between chemotherapeutics and phage K were investigated (illustrated in Fig. 3A). The chosen anticancer drugs showed no obvious influence on the stability of phage K, except Cisplatin, which caused a significant phage titer reduction (> 4 log loss) after 24 h of incubation (Fig. 3B). The impact of these drugs on the propagation efficiency of phage K was also evaluated in S. aureus suspension. In accordance with the parasitic characteristic of phage K, the titer increased by 2 log PFU in the presence of its host compared to that incubating in PBS (Fig. 3C). Irinotecan showed no influence on the replication of phage K. In contrast, Gemcitabine partially inhibited the propagation of phage K (with ~ 1 log reduced). Although no obvious influence on phage stability was detected with Doxorubicin in PBS, phage titer was significantly reduced by the drug when incubated with S. aureus (~ 4 log reduction compared with the drugfree incubation). As for Cisplatin, although it can deactivate phage K, it did not affect phage propagation in the presence of host bacteria that ~ 3 log higher titer was detected comparing with incubation in the absence of host. Drug contents in the presence of phage K were analyzed (Fig. 3D). The four tested drugs showed different levels of content decrease. Relating to clinical use, drug content variation may affect the anti-tumor effect, which is unknown and needs further verification in future work. In addition, the presence of S. aureus paradoxically resulted in drug content reduction (Fig. 3E). In accordance with our findings, Geller et al. also found intratumor M. hyorhinis metabolized gemcitabine into its inactive form, which induced tumor resistance to the drug 21 .
Phage K with chemotherapeutics for combating cellular infection. A coculture model of cancer cells infected with S. aureus was established to further evaluate the intrinsic antimicrobial effect of phage K. Figure 4A displays the time-kill profile of phage K against S. aureus in the coculture. Interestingly, different from the sustained upward trend of OD 600 in the cancer cell-free experiments (Fig. 1), a sharp decline of OD 600 was observed in the coculture after 8 h. Given that no OD 600 values (negligible) were contributed by the seeded cancer cells, the decrease of OD 600 values may be understood as the number of suspended bacteria reduced. To verify this, plate counting was used to recover the viable CFUs of both suspended/planktonic bacteria and settled/cell-associated ones. As shown in Fig. 4B, the cell-associated population was found to be fourfolds higher than that of suspended bacteria in the control group, which implied that S. aureus was more prone to inhabit with cancer cells.
In the coculture experiment, the highest MOI used in this study (10 -1 ) enabled phage K to efficiently destroy the planktonic bacteria and prevent > 3 log CFU of S. aureus from associating to cancer cells (Fig. 4A,B). Likewise, MOI at 10 -2 was also able to combat bacteria in the coculture with a slightly lower efficiency. However, lower MOIs (10 -3 and10 -4 ) showed minimal/unobvious antibacterial effects in the coculture.
The four investigated anticancer drugs showed no significant difference in planktonic and cell-associated bacteria killing. Although the OD 600 value in the Doxorubicin group marginally descended within the first 5 h of incubation, it rebounded with the prolongation of culture time until the end (depicted in Fig. 4C), resulting in only ~ 1 log CFU decrease after 24 h (Fig. 4D). The other drug treatments showed < 1 log CFU reduction.
The antibacterial capacity of phage K against S. aureus was further evaluated in the presence of chemotherapeutics. OD 600 values were quickly reduced within 5 h in the phage K-anticancer drug combinations (Fig. 4E). In accordance with the observed OD 600 trend, planktonic bacteria were considerably killed by the combined therapy. Combinations could also effectively eliminate the cell-associated bacteria, resulting in ~ 4 log CFU decline (Fig. 4F), and the elimination capacity depends on the drug, following an order of Doxorubicin > Gemcitabine > Cisplatin > Irinotecan (Fig. 4F).

Visualization of bacterial infection to cancer cells.
To further elucidate the dynamic interplay of S. aureus with cancer cells, fluorescence microscopy was utilized. Image analysis indicated that S. aureus uniformly distributed in the coculture well during the first 2 h of incubation (Fig. 5). With the extension of infection time, the extent of bacteria associated with the cancer cells was progressively enhanced, with conspicuous bacteria enriched and scattered around the nucleus after 8 h (Figs. 5 and S2). A notable reduction of bacteria numbers was observed with phage K treatment for 2 h. Furthermore, much fewer extracellular bacteria were able to be seen after 8 h of phage K treatment, and a lower level of bacteria internalized into cancer cells was also observed (Fig. 5). Altogether, images revealed that phage K was capable of eliminating the extracellular S. aureus and preventing the cellular infection by these bacteria to a considerable extent. The observation was in accordance with the quantitative data (Fig. 4B).
Similar to the tendency observed in the phage-alone group after 2 h of treatment, combinations also resulted in lower green fluorescence of bacteria compared to the control. And this trend was persistently maintained at 8 h post-infection. Notably, the prevention of cellular infection by S. aureus was remarkably promoted. The  Fig. 6) revealed that Doxorubicin tended to disrupt the cytoskeleton integrity of the cancer cells, which was not observed in S. aureus-infected cells or cells after phage K single treatment.

Bacteria migration under different treatments.
To evaluate the capacity of various treatments on combating S. aureus migration (which may depend on its deformability), a Transwell system was used. Bacteria migrated from the upper insert into the lower chamber through a porous membrane (pore size set at 400 nm) and the migrated quantity was counted. A schematic diagram and obtained data were depicted in Fig. 7A,B.
After 24 h of incubation, 7 log CFU/mL of bacteria migrated. Compared with the bacteria control group, challenge with phage K did not have any effect on the migration potency of S. aureus. A similar tendency was also found in Irinotecan treatment. In contrast, bacteria counts showed a much lower level of S. aureus survival after migration following treatments with Cisplatin (~ 1 log reduction), Gemcitabine (~ 1.5 log reduction), and Doxorubicin (~ 3 log reduction). The concomitantly-applied phage K and anticancer drugs showed a comparable tendency in managing bacterial migration to that found with anticancer drug single use, respectively. www.nature.com/scientificreports/ The migration of phage K through the porous membrane was also evaluated. As shown in Fig. 7C, regardless of the presence of S. aureus, less than 6 log PFU/mL of phages were able to pass through the membrane. This phenomenon may explain the inefficacy of phage K alone use in combating S. aureus migration.

Discussion
Patients immunocompromised by cancer and antineoplastic chemotherapy are susceptible to infections caused by opportunistic bacteria, with S. aureus being the most prevalent single pathogen. To restrict the use of antibiotics and develop novel agents, an increasing number of recent publications highlighted the need for a timely assessment of phage-based solutions 22 . For cancer patients, cancer chemotherapeutics and antibacterial agents are often administered concomitantly, and hence, it is preferable to understand the interactions between these two categories and the overall antibacterial efficacy. www.nature.com/scientificreports/ The impact of chemotherapeutic drugs on the viability of phage K may be derived from their unique mechanisms of action, and consequently, may interfere with the antibacterial outcome. Among the four tested drugs, only Irinotecan showed negligible influence on phage K, therefore, its use with phage K displayed identical bacterial killing as noted in the phage K-alone treatment. Cisplatin declined phage K titer in PBS, and in turn, the antibacterial effect of phage K was weakened. Gemcitabine also showed an impediment of phage K proliferation in S. aureus, but its use with phage K was still proved efficient in destroying S. aureus. On the contrary, Doxorubicin could not allow phage K to attain its effective potency, leading to compromised antibacterial performance. Additionally, it cannot be overlooked that the antibacterial effect of Doxorubicin (Fig. 2B) may play a role in impairing phage K proliferation. Also, one of the mechanisms of action of Doxorubicin on targeting cancer cells and bacteria is via direct membrane damage due to lipid oxidation 23 . We speculated that Doxorubicin may disturb the integrity of S. aureus cell membrane but failed to kill the bacteria. The surface altered bacteria may have limited receptors for phage adsorption and hence their infection.
Data/visualization (Figs. 4B and 5) showed that S. aureus were prone to inhabit with cancer cells. These cancer cells-associated S. aureus can camouflage as a trojan horse to establish an infection reservoir and subsequently become pathogenic within tumors, for a wider dissemination 24,25 . The performance of phage K in preventing bacterial invasion to cancer cells was acceptable (> 3 log CFU decrease). Although specific anticancer drugs displayed interference on phage K performance against S. aureus, this disturbance no longer presented in the coculture. In contrast, these drugs promoted the antibacterial effect of phage K. Among the combinations, phage K with Doxorubicin achieved the strongest antibacterial invasion capability, with 22-folds enhancement in prevention of cellular infection compared with phage K alone use. It is reasonable that Doxorubicin can assist phage K to combat cell-associated S. aureus, since the drug is originally used to deal with cancer cells. In consistence with this description, Doxorubicin was found (via SEM visualization) to disintegrate the compact surface morphology of the cells. The resulting loose and porous morphology may expose relatively-more bacteria for phage infection, and meanwhile, facilitate phage K internalization to target the resided host.
In addition to assessing the performance of treatments on cellular infection prevention, the efficiency to manage bacteria migration was also evaluated. According to a recent review, studies revealed that S. aureus can deform to invade and colonize the narrow confines, which are 100-600 nm in diameter 26 . This characteristic may facilitate antibacterial-and immune-attack evasion of S. aureus and the migration of the pathogen can cause a larger zone of infection in the lung and aggravate the status of pulmonary infections, which was more disastrous in view of the tolerant immunosurveillance in lung cancer patients. Deformation may be a precursor for S. aureus migration, and hence, a Transwell system was utilized equipped with porous membrane (pore size set at 400 nm) for testing S. aureus migration. Along with the visualized diminished cellular infection by S. aureus under phage K-Doxorubicin treatment, the combination group can also combat the bacteria migration, which can play a dual role in managing the persistence and recurrence of pulmonary infections with S. aureus. Contrary to the ability in hampering S. aureus infection to cancer cells, phage K cannot manage bacteria migration, which may be due to its limited transmembrane migration capacity.
The pathogenicity of S. aureus involves its ability to express antibiotic-resistant determinants, to implement intracellular infection as well as consecutively invading lung tissues for persistence 2,27 . Pulmonary S. aureus infections are also considered as severe secondary infections with reduced effect of antibiotics against them, especially during the recent years of coronavirus disease 2019 (COVID-19) pandemic 28 . The continuous attack  29 , and prescriptions of broad-spectrum antibiotics, even at high doses, usually lead to negligible effect due to the high frequency of resistance and insufficient intracellular diffusion of antibiotics 28,30 . In this setting, the combination strategy of phage K + Doxorubicin proposed in this study can be also explored in handling secondary S. aureus infections in the post COVID-19 era.

Conclusion
In this study, the feasibility of applying phage K combined with chemo drugs for the management of invasive S. aureus infections in lung cancer was evaluated (Fig. 8). Cisplatin was found to directly reduce phage titers while Gemcitabine/Doxorubicin partially repressed its propagation. In the in vitro coculture model, phage K was proved to effectively destroy the planktonic bacteria but limitedly resist S. aureus invasion to cancer cells. This issue was largely remedied by a combination use with Doxorubicin, which was owing to not only the antibacterial effect of the drug but also the devastation to the cancer cells. In addition, Doxorubicin was able to inhibit S. aureus migration whereas phage K alone failed. Given that cancer chemotherapeutics and antibacterial agents are often administered concomitantly, our work may benefit the future clinical administration options. In this case, we would recommend that Doxorubicin can be used as an effective adjuvant for phage K in the treatment of S. aureus infections developed in cancer patients.  A fresh culture of S. aureus grown from a single colony was incubated overnight in NB at 37 °C under shaking. Then, the bacterial suspension was 100-fold diluted with NB medium/DMEM cell culture medium and incubated until the optical density at 600 nm (OD 600 ) reached 0.6 (~ 10 8 colony-forming unit (CFU)/mL) for further use.

Materials
High titer phage K lysate was produced and collected using well-established protocols 31,32 . The phage K lysate was then purified by anion-exchange chromatography using a CIMmultus QA 1 mL Monolithic Column (BIA Separations, Slovenia). The phage K elution was dialyzed with Phosphate buffered saline (PBS) and the obtained phage K titer was 1 × 10 9 plaque-forming unit (PFU)/mL. A549 lung cancer cells were incubated in DMEM supplemented with 10% FBS at 37 °C with 5% CO 2 .

MIC determination. The minimal inhibitory concentrations (MIC) of the anticancer drugs to the S. aureus
strain were determined using the broth dilution method. Briefly, a range of drug solutions in PBS with twofold serial dilution was prepared and incubated with S. aureus suspension (100-fold dilution of bacteria stock OD 600 = 0.6 in NB) for 24 h. MIC was defined as the lowest drug concentration that prevented bacterial growth, which was determined by OD 600 measurement using a microplate reader (CLARIOstar, BMG Labtech, Germany).
In vitro bacterial killing assays against S. aureus. The antimicrobial activity of phage K/anticancer drugs on S. aureus should ideally be measured by more than one activity assay since results from different assays can vary on a quantitative level and each assay may be biased toward different properties. In this setting, two different in vitro activity assessments, OD 600 reduction and plate colony counting, were adopted here to quantitatively evaluate the killing efficiency of phage K/anticancer drugs. Briefly, the assay was performed in a 96-well plate where S. aureus and phage K were added at various multiplicity of infections (MOIs). Anticancer drugs were used at a predetermined concentration (10 μM), to better compare their performances in bacterial killing and related properties under identical drug molecule amounts. The mixtures were incubated at 37 ℃ without shaking for a total of 24 h. Plates at specific time points were withdrawn for OD 600 measurement, and an aliquot of the bacterial culture (10 μL) was taken for bacterial counting assay after 24 h. www.nature.com/scientificreports/ Phage K titer determination. The impacts of tested anticancer drugs on the viability of phage K were investigated. Briefly, phage K and anticancer drugs at the predetermined concentration (10 μM) were incubated together in either PBS or S. aureus suspension in a 96-well plate at 37 °C. After 24 h of incubation, mixtures were collected from each well and plaque assay was performed to determine phage titer variation. For phage incubated with drugs in PBS, samples withdrawn from the wells were directly diluted for plaque assay. For co-incubation in the presence with S. aureus, samples were centrifuged to removing bacteria before diluted for plaque assay.

Drug content determination.
To evaluate the impact of phage K and S. aureus on drug content, drug concentrations after incubation were measured. Briefly, anticancer drugs at the predetermined concentration (10 μM) were incubated in PBS or S. aureus suspension with/without phage K in a 96-well plate at 37 °C for a specific time interval. Then the mixtures were centrifuged at 8000×g (Thermo Fisher) for 5 min. The supernatant was then processed with acetonitrile to precipitate protein and the obtained solutions were collected for drug concentration quantification through respective analytical methods (HPLC/microplate reader). The variation in drug content (%) was determined as the relative ratio to the drug contents obtained in the PBS controls.
Antimicrobial effect assays in a S. aureus-infected cell culture model. To determine the ability of phage K in preventing cell infection by infectious bacteria, the antibacterial effect assay of phage K was performed in the presence of lung cancer cells. A549 cancer cells were seeded in a 96-well plate at a density of 1 × 10 4 per well in DMEM medium and incubated overnight. After removing the spent medium, the cells were infected with S. aureus suspension at a bacteria-to-A549 cell MOI of 1000 to 1. Afterward, phage K was added into the coculture system at various phage-to-bacteria MOIs (10 -1 to 10 -4 ). Controls without S. aureus infection and phage K treatment were included in the assay. After 24 h of incubation at 37 °C and 5% CO 2 , the upper suspended medium was removed and plated to evaluate the planktonic bacteria counts. In the meanwhile, the A549 cell layer was washed with ice-cold PBS three times, and EDTA-PBS was used to detach the adherent cells.
The cell suspension was subsequently collected and serially diluted with PBS, followed by plating on NB agar for bacterial counting. Similar procedures were conducted to test the antibacterial efficacy of anticancer drugs to both the planktonic and cell-associated bacteria, with the drug concentration fixed at 10 μM for each tested drug.
To determine the antibacterial efficiency of phage K with chemotherapy in lung cancer, bacterial killing assays were performed in the infected coculture system as described above, with the following modifications. Various antitumor drugs would be respectively added into the infected coculture system together with phage K (MOI of phage K to bacteria 10 -1 ). After 24 h of incubation at 37 °C with 5% CO 2 , viable CFUs of both planktonic and cell-associated bacteria were determined by plate counting. Log reductions caused by the different combination treatment groups were evaluated and compared.
Fluorescence microscopy. To qualitatively observe the antibacterial effect of various treatments and the cell infection by bacteria, a fluorescence microscope (Nikon, TI-DH, Japan) was applied. For visualization of antibacterial effect, S. aureus suspension (described above) was challenged with each treatment group for 24 h, and live/dead Bacterial Viability Kit (Thermo Fisher, USA) was used to fluorescently label the treated bacteria. After staining, samples were washed with PBS to remove the left dye and images were captured by fluorescent microscopy.
For visualization of cell infection by S. aureus, bacteria were stained with a green fluorescent dye (SYTO™ 9) while Hoechst blue fluorescent dye was used to stain the nucleus of the cancer cells. A549 cancer cells were seeded in a 24-well plate at a density of 1 × 10 5 per well and incubated in DMEM medium overnight. Thereafter, S. aureus suspension was used for infection of the cultured cells at an MOI of 100 to 1. Therapeutic agents were then added to the coculture system and the mixtures were incubated at 37 °C with 5% CO 2 . At specific time points, the upper suspended media were removed and the cells were washed with PBS for three times. After staining the bacteria and cancer cells with specific dyes, the samples were re-washed with PBS to remove the left dye and the images in the visual field were randomly recorded by fluorescent microscopy.

Scanning electron microscopy (SEM).
To observe the morphology of S. aureus and A549 cancer cells, SEM was used. Briefly, cancer cells were seeded and infected with S. aureus as above-described. After various treatments, colonized cells were fixed in 2.5% glutaraldehyde and dehydrated using ethanol with a series of concentration gradients (50%, 75%, 100%). Thereafter, suspensions were centrifuged and the obtained pellets were collected and visualized by SEM (Quanta 400F FEI).

Migration assay of S. aureus.
To determine the migration of S. aureus and management of various treatments on the migrated S. aureus, a Transwell system was used, equipped with 6.5 mm inserts covered by 0.4 μm polyester membrane. Briefly, the prepared bacteria stock (described above), challenged with different therapeutic options, were applied to the upper inserts (200 μL) while PBS was placed into the lower receiver chambers (1 mL). The migration was carried out at 37 °C for 24 h. After incubation, bacteria transferred into the lower receiver chamber were quantified by plate counting. Counts of bacteria that migrated from the upper inserts to the lower chamber through the membrane were calculated and compared between different treatments. Statistical analysis. All experiments were performed at least in biological and technical triplicates. Data were analyzed using GraphPad Prism 9 software. All experimental data were expressed as mean ± standard deviation (SD). Whenever appropriate, comparisons among multiple groups were performed by one-way analysis www.nature.com/scientificreports/ of variance (ANOVA), while a two-tailed Student's t-test was conducted to identify the statistical differences between the two groups. A probability value (p) of less than 0.05 (p < 0.05) was considered to be statistically significant.

Data availability
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.