All the work related to cell line and bacteria were performed aseptically, under laminar airflow. Clinical strain RS-307 of A. baumannii was grown in Luria-Bertani broth (LB) media at 37°C till it reached optical density to 0.6 OD. After sufficient bacterial growth, streaking was done on the Luria agar plate to isolate the single bacterial colony. Single colony was used for suspension culture. Prior to coinfection, bacterial culture was centrifuged and re-suspended in PBS to remove deleterious effects of toxins (cytotoxins) present in the culture.

The human alveolar basal epithelial cell line A-549 (purchased from Cell repository, NCCS, Pune) was cultured in high glucose Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% heat inactivated fetal bovine serum albumin (heat inactivation was performed at 56°C for 30 min), amphotericin B (2.5 µg/ml), vancomycin (50 µg/ml), gentamicin (50 µg/ml), and 1% HEPES in a humidified, 5% CO₂ at 37°C. Passage of A-549 human pulmonary cell line in 25 cm² flask was done after every 2–3 days intervals when they were confluent. The cells were seeded for 24 h in 96 well plates for MTT and LDH assay and 6-well plates for coinfection with the A. baumannii.

Confluent A-549 cells flask was washed twice with PBS buffer (pre-warm at 37°C). The cells were detached with the help of 0.05% trypsin-EDTA and complete media was added. The suspension was centrifuged and cells pellet were re-suspended in the fresh DMEM medium containing 10% fetal bovine serum without antibiotics, at a concentration of 2 × 10⁵ cells/ml. Simultaneously, bacterial culture was inoculated in 100 ml of LB broth and was grown at 37°C with vigorous shaking (180 rpm) till optical density reached to 0.6. We have counted colony-forming unit (CFU) of the bacteria before coinfection.

A-549 cells were grown at 37°C in 5%CO₂ till they reached at least 90% confluency without contamination. The cells were washed with pre-warm PBS. 1 ml of fresh medium supplemented with 10% serum without antibiotics was added to each well. The control wells were prepared by adding fresh medium without A-549 cells in three wells. These wells were used as a blank for addition of bacterial strains. Coinfection was achieved by adding an aliquot of each bacterial culture to each well containing A-549 cells (in triplicate) and in the blank. This coinfection represents MOI (multiplicity of infection) of 3:1 (bacteria, 1.2 × 10⁷: cells, 0.4 × 10⁷). Infected A-549 cells were incubated in incubator for 24 h at 37°C with 5% CO₂.

We have counted bacterial populations of four different stages of infection, i.e., before coinfection, remain non-interact during coinfection (suspension culture in DMEM media which contain the free bacteria), interacting or adhered bacteria on the surface of A-549 (which were isolated after mixing of cell pellet in the PBS and centrifugation, as explained in next section) and internalized bacteria into A-549 (which were isolated after homogenization). The CFU was determined for above four stages.

After 24 h incubation, washing was done 3 times with pre-warm PBS. 1 ml of PBS (containing 0.1 mM protease inhibitor PMSF) was added. Adhered bacteria were detached and centrifuged at 3000g for 5 min to separate the loosely adhered bacteria. The pellet was dissolved in 1 ml PBS (with 0.1 mM PMSF). The dissolved pellet was homogenized for 1 min. Homogenized sample were centrifuged at 7000g for 10 min. The supernatant was stored at −80°C containing A-549 cell lines proteins. The pellet (containing internalized bacteria) was suspended in 1 ml PBS with 0.1 mM PMSF. This suspended pellet was sonicated for three cycles of 30 s (20 kHz, 130 W) with the interval of 1 min. The sonicated samples were centrifuged at 10000g for 10 min. The supernatant contained bacterial protein and was stored at −80°C. The samples were collected before homogenization and sonication, and used for bacterial CFU counting that represent adhered and internalized bacteria, respectively. Protein concentrations were measured for bacterial samples (internalized and adhere) and cell line (before and after bacterial infection) using Bradford methods.

Polyvinylpyrrolidone-capped AgNPs were prepared as per our published methods (12). The size and zeta-potential of synthesized PVP-capped AgNPs were monitored using dynamic light scattering-based particle analyzer. In vitro susceptibility of PVP-AgNPs (30 µM) was performed on carbapenem-resistant strain of A. baumannii using disc diffusion assay and growth kinetics analysis as per our published protocol (18).

For the treatment, four T-25 flasks were prepared. First flask was used as a control for cell line, second flask was used for bacterial infection, third flask was for AgNPs treatment after bacterial infection, and fourth flask as negative control for AgNPs treatment. All four flasks were incubated at 37°C for 24 h with 5% CO₂. Coinfection was achieved by adding 1.2 × 10⁷ CFU that represents MOI of 3:1 (bacteria, 1.2 × 10⁷: cells, 0.4 × 10⁷). Infected A-549 cells were incubated in the CO₂ incubator at 37°C for 24 h with 5% CO₂. 500 µl of culture media was taken for counting non-interacting bacteria. The media was discarded gently and flasks were washed mildly two times by 2 ml PBS. All the washed cell lines were collected in 1 ml PBS. This suspended cell line mixture contains surface interacting bacteria and cell line (with or without engulf bacteria). This mixture was centrifuged at 3000g for 5 min and supernatant was used for CFU counting of interacting bacteria. The collected samples were homogenized for 3 min (1 min pulse each) and centrifuged at 7,000g for 10 min. The supernatant contains cell line proteins and pellet contains engulfed bacteria. The pellet was dissolved in 500 μl PBS and used for CFU counting of engulf bacteria. Suspended bacteria was further sonicated and centrifuged. The supernatant contained bacterial proteins. The cell line proteins and bacterial proteins of different conditions were used for the SDS-PAGE analysis.

Cytotoxic effect of PVP-capped AgNPs (15–300 µM) on the A-549 cell line was investigated using MTT (EZcount™ MTT Cell Assay Kit) and LDH assay (EZcount™ Lactate Dehydrogenase Cell Assay Kit) to find the non-toxic dose as well as IC₅₀ concentration of PVP-AgNPs.

Viability of A. baumannii in infected A-549 cell line was monitored using MTT assay. The triplicated experiment was performed on A-549 alone, A-549 cell line infected with A. baumannii, and A. baumannii-infected A-549 cell line treated with PVP-AgNPs. The viable internalized bacteria were calculated by subtracting total viability of infected A-549 cell line with A. baumannii by the viability of uninfected A-549 cell line.

Reactive oxygen species production was monitored by nitroblue tetrazolium (NBT) reduction assay for uninfected A-549 cell lines, A. baumannii infected A-549 cell lines (A-549), and A. baumannii-infected cell lines after PVP-AgNPs treatment. NBT stock solution (10 mg/ml) was prepared in distilled water and kept in dark at 4°C. A working NBT solution (0.3%) was prepared in culture medium just before utilization. It was diluted three times in each well, giving a final concentration of 0.1% NBT in the well. 50 µl supernatants were eliminated when the cells were adhered and replaced it by 50 µl of 0.3% NBT working solution. After 2 h of incubation, supernatant was removed and cells were fixed by adding of 200 µl of methanol and washed twice with 70% methanol, then dried. The formazan deposits were solubilized in 120 µl 2 M KOH and 140 µl DMSO. After vigorous mixing of all the contents of wells, absorbance was recorded at 620 nm using UV-Vis spectrophotometer. To remove the contribution coming from the ROS produced by A. baumannii, a bacterial control was also taken.

After infection, A-549 cell lines were incubated for 24 h and numbers (CFU) of A. baumannii were monitored for interacting or adhered bacteria on the surface of A-549 and internalized bacteria into A-549. The CFU result is presented in the Table 1. Table shows that about 41% bacteria remain un-interacted, 39.6% bacteria adhered to the surface of A-549 cells, and 18.4% bacteria were engulfed or internalized. The CFU results of these stages are presented in the Figure 1. The infection of A-549 cell line with the A. baumannii leads to the lysis of the A-549 cell lines and also changes the morphology of A-549 cell lines that can be seen in Figure 2.

Synthesized PVP-capped AgNPs have maximum absorbance peak around 400 nm for AgNPs (Figure 3A) and size of 133 nm with polydispersity index of 23% (Figure 3B). Similarly, zeta-potential of nanoparticle was found to be −34 mV. Growth kinetics of RS-307 strain of A. baumannii was determined in the absence and presence (30 µM) of PVP-capped AgNPs. Optical density was measured at 605 nm at an interval of 30 min using UV-Vis spectrophotometer. Growth curves were prepared using absorption data. The experiment was performed in triplets for untreated and treated samples and average value was used to prepare the plot. Relative growth curves were prepared for comparison purpose. The result showed that PVP-capped AgNPs showed good antimicrobial activity against carbapenem-resistant strain of A. baumannii (Figure 3C).

The cytotoxic effect of PVP-capped AgNPs on the human pulmonary epithelial cell lines (A-549) was identified using MTT Assay and LDH assay. Figure 4 showed that the concentration at which it showed bactericidal activity has no cytotoxicity. The IC₅₀ value for PVP-AgNPs against A-549 cell line was found to be 130 µM, which is four times higher than the concentration showing antimicrobial activity. The 30 µM PVP-AgNPs also does not have hemolytic activity.

The infected cell line of A-549 was treated with PVP-AgNPs. The infection leads to the death of A-549 cell line (Figure 5B) but PVP-AgNPs protects pulmonary cell line (A-549) from the infection of A. baumannii (Figure 5C). The morphologies of PVP-capped AgNPs-treated A-549 cell line (Figure 5C) are very similar to the uninfected cell lines (Figure 5A). Therefore, it can be suggested that presence of PVP-AgNPs prevent infection of carbapenem-resistant strain of A. baumannii in A549 cell line.

Colony-forming unit counting in the absence and presence of PVP-AgNPs showed that PVP-AgNPs reduce the number of surface attached bacteria from about 40 to 2.2%. The number of intracellular bacteria was also reduced from 18 to the 1.25%. Therefore, we can say that PVP-AgNPs reduces the attachment and internalization of A. baumannii on the pulmonary cell model.

Intracellular viability was calculated by subtracting absorbance of infected A-549 cell line with absorbance of uninfected A-549 cell lines. The result of viability test showed that there is 80% decrease in the viability of intracellular A. baumannii after treatment (OD 0.12) with 30 µM PVP-AgNPs as compared to untreated sample (OD 0.56). It also enlightened that PVP-AgNPs not only inhibit free A. baumannii but also inhibit the intracellular A. baumannii.

Reactive oxygen species are used by pulmonary epithelial cells against pathogens (19). ROS can kill pathogens (such as A. baumannii) directly by causing oxidative damage to biomolecules or indirectly by stimulating pathogen elimination by various non-oxidative mechanisms (19). Therefore, we have monitored the ROS production during the infection as well as treatment. Result (Figure 6) showed that A. baumannii infection promotes threefold increase (OD at 620 nm is 0.83 for infected and 0.27 for uninfected) in the production of ROS. This is correlated with the previous result, which showed that ROS production increases in pulmonary epithelial cells after bacterial infection. The treatment of PVP-AgNPs shifted the ROS level to normal, which also confirms that approximately all the intracellular bacteria were killed by PVP-AgNPs. The result is correlated with the bacterial viability assay, which also confirms that about 80% intracellular bacteria were killed by the PVP-AgNPs.

The estimated value of protein isolated from suspension bacteria, internalized bacteria, cell line blank, and coinfected cell line were found to be 4.20, 1.51, 2.0, and 1.77 µg/µl, respectively. The SDS-PAGE profiling of RS-307 bacterial proteins (Figure 7A) and A-549 cell line proteins (Figure 7B) in the presence and absence of each other. The result confirms that after coinfection, protein with 43 kDa overexpressed while a protein with 35 kDa undergoes downregulation in the protein profiling of bacteria (Figure 7A). Similarly after infection a protein about 40 kDa expression is markedly decreased and banding pattern of A549 is totally different after Acinetobacter infection.