Doxorubicin-loaded bacterial ghosts of Streptococcus pneumoniae and Klebsiella pneumoniae: An In-vitro in-silico lung cancer cells study

2.1 Preparation of S. pneumoniae and K. pneumonia BGs

We prepared S. pneumoniae and K. pneumonia BGs by applying the previously invented protocol (16). Briefly, the fresh cultures of each bacterial strain were inoculated in 7% v/v Tween-80 Muller-Hinton broth and kept incubated for one day at 37°C. Every culture was treated with an organic acid (lactic acid, pH=3.6). The batches of BGs were prepared by lyophilization and stored at room temperature (25°C).

2.2 Confirmation of well-characterized BGs’ formation

The ghost’s formation was confirmed by visualization of cell intactness and hole formation, followed by a light and scanning electron microscope. The quantification of released proteins and DNA from the settled cells, and cell death (checked by subculturing and sterility testing) confirmed their cellular evacuation (Rabea et al., 2018).

2.3 Loading of S. pneumoniae and K. pneumonia BGs with doxorubicin

Loading of lyophilized ghosts with DOX (LC Laboratories, Woburn, MA, USA) was optimized to reach the maximum loading capacity according to the standard protocol (Paukner et al., 2004). Four cofounding factors were included: incubation temperature (25°C), contact time (10 minutes), drug concentration (1 mg/ml), and medium buffer at pH 7. The weight of lyophilized BG (10 mg) was constant in all trials. In the beginning, four trials were performed. In each trial, three factors were kept constant, and the fourth one was variated. The first trial investigated the effect of different temperatures (5°C, 30°C, 40°C, 50°C, and 60°C). The second trial evaluated the impact of modifying contact time (10, 20, 30, 40, and 60 minutes). The third trial tested the effect of different drug concentrations (1, 2, 3, 4, and 6 mg). The last trial investigated the effect of changing the medium’s pH (6,7, 8, 9, and 10).

After centrifugation at 5000 x g for 10 minutes, the obtained loaded BGs with DOX were washed four times with Tris HCl buffer and then lyophilized. In all experimental conditions, DOX stability has been tested.

2.4 Loading and encapsulation capacities calculation

The loading capacity was calculated according to the following equation:

$\text{Loading capacity }(\text{LC)}=\frac{\text{Entrapped DOX}}{\text{ BG weight }}(\mu \text{g/mg}).$

The encapsulation capacity was calculated according to the following equation:

$\text{Encapsulation Capacity (EC }\%\text{)}=\frac{\text{Entrapped DOX }}{\text{Total added DOX }} \%(\mu \text{g/}\mu \text{g}) .$

2.5 Determination of particle size and zeta-potential of DOX. loaded and free BGs particles

Particle size and zeta-potential of SPG, KPG, SPG-DOX, and KPG-DOX. were measued using Zetasizer Nano ZS 90 device (Malvern, UK), with Nano laser at 633 nm equipped with Helium–Neon laser (633 nm) as a light source, with the detection at 90 degree scattering angle at room temperature (25°C) (Clogston and Vermilya, 2020).

2.6 Evaluation of the loaded DOX quantity

To quantify the loaded DOX, the ghost’s matrix was disrupted by adding 2 ml of 5% sodium lauryl sulphate (SDS) solution and was kept at 65°C, followed by adding 1 ml of 10 mM sulfuric acid. After extra incubation for 10 minutes, the supernatant resulting from centrifugation was analyzed by reversed-phase high-performance chromatography (RP-HPLC) method to quantify the released DOX as previously reported (19). The loaded released DOX was detected by the previously developed system using a Waters HPLC system (Waters, USA) attached with a 1515 isocratic HPLC pump, 717 plus autosampler with injection volume of 10 μL, quaternary LC-10A VP pumps, and a programmable 2487 dual λ absorbance UV-Visible detector. At 25°C, a Nucleodur 150 mm × 4.6 mm RP C8 column filled with 5 μm filler was used as the stationary phase. The mobile phase consisted of ethyl acetate and methanol (7:3% v/v). The elution was carried out at a flow rate of 1.0 mL/min with UV detection at 254 nm. The procedure of loaded quantities analysis (RP-HPLC) was evaluated for linearity, accuracy, precision, robustness, and selectivity (Alsuwyeh, 2017). Linearity was validated by constructing a calibration curve using varying dilutions. Accuracy and precision were validated using the standard addition method and the intraday and intermediate precision determination methods, respectively. Robustness was determined by comparing the changes observed in the chromatographic responses when the concentration was increased or decreased by one step. Selectivity was determined by injecting different concentrations of DOX. (Alsuwyeh et al., 2018).

2.7 Drug release monitoring and kinetic modeling

Three DOX formulations: Free DOX, SPG-DOX, and KPG-DOX were monitored for in vitro release by the drug release monitoring method, dynamic dialysis, following the standard protocol (Ge et al., 2018). One ml (4 mg/ml) DOX was involved in each formula and kept in the dialysis membrane bags (MWCO 12000 g/mol (6-8 kDa), Livingstone, NSW, Australia). The dialysis bag (20 ml receptor volume) was soaked in distilled water for 12 hrs. before its use. The bags were incubated in phosphate buffer saline (PBS) at 37 C° $\pm$ 1 and pH 7.4. The shaking speed was slow and adjusted to 50 rpm. From each DOX formula, 1 ml was withdrawn for analysis by the previously mentioned RP-HPLC protocol at the following time intervals: 5, 10, 15, 30, 60, 120, 180, 240, 360, 720, 1440, 2880, 4320, 5760, and 7200 minutes. To maintain sink condition, 1 ml of fresh PBS was added after each withdrawal step. The calculated percentages of the released DOX versus time were plotted to obtain the drug release curve and acquire the linear regression equations (Modi and Anderson, 2013). The most-fit mathematical model was selected to represent the release kinetics (Dash et al., 2010).

2.8 Determination of antiproliferative activity on two different lung cancer and normal cell lines.

To determine the effect of BGs on the viability of lung cancer cells using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay, H69A (ATCC NCI-H69) small cell lung carcinoma, A549 (ATCC CRM-CCL-185) lung adenocarcinoma cancer cell line, and immortalized human embryonic kidney cell line HEK-293 (ATCC CRL-1573) cells were purchased from the American type culture collection (ATCC, Rockville, MD, USA). HEK-293 cells are used as normal cells. The cells were cultured and maintained following the ATCC protocols using standard aseptic techniques in Dulbecco’s modified eagle’s medium (DMEM), supplemented with 10% fetal bovine serum (FBS), 100 UI/mL penicillin G, and 100 µg/mL streptomycin, then incubated at 37°C in 5% carbon dioxide/95% air.

The cytotoxicity of five formulae: Free-DOX, KPG, SPG, KPG-DOX, and SPG-DOX was evaluated on H69A, HEK-293, and A549 cells. The cells were seeded in a 96-well culture plate at a density of 1 × 10^4 cells per well in 100 µL of culture medium and incubated for 24 h. A serial dilution of the five formulae at concentrations ranging from 100, 50, 25, 12, 6, 3 to 1 μg/ml was prepared. The BG suspension was dispersed by sonication for 3 min, then incubated with cells for 24 h. A total of 20 µL of 2.5 mg/ml of MTT solution in PBS was added to the cells, which were further incubated for 4 h. at 37°C. The MTT solution was removed entirely, and 100 µL of dimethyl sulphoxide (DMSO) was added to each well to solubilize the formazan crystals. Absorbance was measured at 540 nm as mentioned (Ashour et al., 2014).

2.9 Analysis of cells apoptosis by flow cytometry

The apoptotic effect for Free-DOX, KPG, SPG, KPG-DOX, and SPG-DOX was evaluated against H69A, A594, and HEK-293 cells using flowcytometry. Cells were treated with double IC₅₀ (nmol/L) for 24, 48, and 72 h. Cells were then harvested, suspended in 500 µL Annexin binding buffer (ABB), and incubated with 5 µL/mL Annexin-APC (Annexin-V conjugated to allophycocyanin (Life Technologies, Grand Island, NY, USA) for 15 min. Treated samples and controls were analyzed using FACSCanto-II (BD-BioSciences, San Jose, CA, USA) as reported (Ashour et al., 2014).

2.10 Protein-protein docking study

3.1 Confirmation of well-characterized preparation of BGs from S. pneumoniae and K. pneumoniae

3.1.2 Visualization of the created holes using a scanning electron microscope (SEM)

The prepared ghosts’ samples showed dozens of perforated cells, as shown in the scanning electron micrograph Figs. 1(a-f).

Close

Fig. 1. Scanning electron micrograph showing surface tunnels through the cellular walls of S. pneumoniae ghosts. (a): X11,000; (b): X17,000; (c): X30,000, and K. pneumoniae ghosts. (d): X9,500; (e): X17,000; (f): X27,000.

Export to PPT

3.2 The loading of BGs with doxorubicin

The loading conditions were temperature, pH, loaded quantity, and ghost-drug contact time. The optimized loading conditions were 50°C, pH 9, 1 mg DOX, and 10 minutes of contact time. These conditions were selected as they achieved the maximum loading capacities (31.2 ±3.1, 28.8 ± 2.6 ug Dox/mg ghost) (n=3) for SPG and KPG, respectively, and maximum encapsulation capacities (31.2% ± 3.1, 28.8% ± 2.6) (n=3) for SPG and KPG, respectively.

3.3 Determination of particle size and Zeta-potential of DOX. loaded and free BG particles

The particle sizes of different ghosts, either empty or loaded, were slightly varied. They were 1.11±0.2, 1.26±0.1, 1.34±0.1, 1.34±0.2 μm (n=3) for SPG, SPG-DOX, KPG, and KPG-DOX, respectively. Zeta-potentials of different formulae were also slightly varied: -27.05±0.8, 26.5±0.3, 33.7±1.2, 32.3±0.9 mV(n=3), for SPG, SPG-DOX, KPG, and KPG-DOX, respectively.

3.4 Drug release monitoring and kinetics modeling

The loaded DOX on the two different BGs showed a different release pattern than the Free DOX. The loaded DOX was released from the BGs matrices of S. pneumoniae and K. pneumoniae ghosts in a pattern obeying the Higuchi model kinetic. The plot of the cumulative percentage of the released drug versus the square root of time showed the best fit, with the highest R² (0.9778 and 0.973, in SPG and KPG, respectively), as shown in Fig. S2a. By applying other kinetic models and comparing their R-squares, the applied Higuchi model was the best fit, having the largest R². After 5 minutes of starting the experiment, 47.8% and 43.9% of the loaded drug were released from SPG and KPG, respectively. At the end of the experiment (5 days), 59.5% and 54.9% of DOX were released from SPG and KPG, respectively. The Free DOX was released from the dialysis membrane bag in a pattern following the first-order kinetic model. The log of the cumulative percentage of drug remaining versus time plot exhibits the best fit with the highest R² (0.8705), as represented in Fig. S2b. The percentages of the released free drug were 41.6%, 47.2%, 94%, and 99.9% after 5 min, 6 hrs., 12 hrs., and 5 days, respectively (Fig. S2b).

3.5 Antiproliferative effect of BGs on lung cancer cells

The percentage of cell viability of multidrug-resistant (MDR), small cell lung carcinoma (SCLC), cell line (H69A) represents antiproliferative activity. The tested drugs at different concentrations of 1, 3, 6, 12, 25, 50, and 100 μg/ml, showed concentration-dependent antiproliferative activity. In comparison to Free-DOX, achieving an average of 46.6% SCLC viability, the SPG-DOX resulted in an average of 10.8% SCLC viability Fig. 2. The difference was significant (P-value ˂ 0.001, One-way 2analysis of variance (ANOVA) post hoc Tukey’s test) within 24, 48, and 72 h. The calculated IC₅₀ value of SPG-DOX was 4.1 μg/ml (95% CI 3.1-5.3), while the IC₅₀ value of Free-DOX was 27.2 μg/ml (95% CI 13.9- 52.9), Table 1. The SPG also accomplished better antiproliferative activity (12.6%) than Free-DOX using the same concentration as depicted in Fig. 2 (P-value ˂ 0.001, One-way ANOVA post hoc Tukey’s test). The calculated IC₅₀ of SPG was 9.9 μg/ml (95% 4.9-19.5), Table 1. Both KPG and KPG-DOX achieved 28.8% and 37.7% viability, respectively (Fig. 2). The IC₅₀ of KPG was 27.5 μg/ml (95% CI 19.3-39.3), while the IC₅₀ of KPG-DOX was 21 μg/ml (95% CI 14.6-30.5), Table 1.

Close

Fig. 2. Antiproliferative activity of different formulae on the H69A. The activity was represented by plotting the viability percentage versus different concentrations (1, 3, 6, 12, 25, 50, and 100 μg/ml) of drugs at three different times (a): 24, (b): 48, and (c): 72 hrs. Free-DOX, free doxorubicin; KPG, K. pneumoniae ghosts; SPG, S. pneumoniae ghosts; KPG-DOX, Doxorubicin loaded K. pneumoniae ghosts; SPG-DOX, Doxorubicin loaded S. pneumoniae ghosts.

Export to PPT

Exploring the antiproliferative activity of different tested groups shows different percentages of viability of A549 cells. Like the effect pattern of the tested drugs on H69A, the cells showed dose-related antiproliferative activity at various concentrations. The Free-DOX achieved an average of 38.5% cell viability, while the SPG-DOX reached an average of 23.1% cell viability (Fig. 3). The difference was significant (P-value ˂ 0.01, One-way ANOVA post hoc Tukey’s test). The calculated IC₅₀ of SPG-DOX was 4 μg/ml (95% CI 2.1-7.6), while the IC₅₀ of Free-DOX was 7.9 μg/ml (95% CI 3.3- 18.8), Table 1.

Close

Fig. 3. Antiproliferative activity of different formulae on A549 cells. The activity was represented by plotting the percentage of viability versus different concentrations of drugs at three different time points (a): 24, (b): 48, and (c): 72 hrs..

Export to PPT

The KPG-DOX also achieved better antiproliferative activity (26.7% viability) than Free-DOX using the same concentrations as shown in Fig. 3 (P-value ˂ 0.01, One-way ANOVA post hoc Tukey’s test). The calculated IC₅₀ of KPG-DOX was 43.6 μg/ml (95% CI 34.3-55.8), Table 1. Both KPG and SPG achieved 33.1% and 45.3% viability, respectively Fig. 3. The IC₅₀ of KPG was 66 μg/ml (95% CI 19.3-39.3), while the IC₅₀ of SPG was 46 μg/ml (95% CI 14.6-30.5), Table 1.

The lowest toxicity of all tested formulae on HEK 293 was observed after 24 hrs. of treatment, showing more than 60% cell viability Fig. 4. The safest drug formula was the SPG-DOX group since the HEK 293 cells showed more than 70% viability. Additionally, SPG-DOX IC₅₀ was 155.3 μg/ml, while the IC₅₀ of Free-DOX was 121.5 μg/ml (Table 1. Compared to the toxicity of free (viability percentage 65.1%), SPG-DOX and KPG-DOX (viability% 70.3 and 66.3, respectively) were safer (P-value >0.001, Two-way ANOVA post hoc Tukey’s test). The KPG toxicity was almost equal to that of the free drug. At the same contact time (24 h), the SPG was more toxic than the free drug, showing 59.1% viability (P-value >0.001, Two-way ANOVA post hoc Tukey’s test) and the IC₅₀ equal to 88.4 μg/ml, Table 1.

Close

Fig. 4. Antiproliferative activity of different formulae on HEK 293 cells. The activity was represented by plotting the percentage of cell viability versus different concentrations of drugs at three different time points (a): 24, (b): 48, and (c): 72 hrs.

Export to PPT

After 48 h. of treatment, the tested groups achieved more than 55% viability, but the KPG-DOX was the safest (viability % was 68.3%). After 72 h, the tested groups showed more than 50% viability. The safest formula was also KPG-DOX with a viability percentage of 64.3%, with IC₅₀ equal to 116 μg/ml (P-value > 0.001, Two-way ANOVA post hoc Tukey’s test) as shown in Table 1. The more toxic formulae were SPG, SPG-DOX, and free-DOX groups, resulting in 51.4%, 51.6%, and 55.4% viability, respectively, Fig. 4.

3.6 Effect of BGs on inducing cell apoptosis by flow cytometry

The effect of the tested drug formulae showed wide variations in the cell cycle phases of treated cells. There was no significant difference between Free-DOX, SPG, and control groups (P value <0.05, Two-way ANOVA post hoc Tukey’s test), resulting in less than 20% apoptosis in H69A cells (Fig. S3). The most effective drug formula was KPG-DOX (58.2% apoptosis), followed by KPG (51% apoptosis). The SPG-DOX achieved 31.5% apoptosis, (P-value >0.001, Two-way ANOVA post hoc Tukey’s test), as presented in Supp. Fig S3.

The apoptotic effect on A549 cells was the least in the control and KPG groups (less than 20% apoptosis). The highest apoptotic activity (57.7%) was achieved using the SPG-DOX group as compared to the control group (P value > 0.001, Two-way ANOVA post hoc Tukey’s test) (Fig. S3). In A549 cells, KPG-DOX and SPG achieved 44.1% and 37.2% apoptosis, respectively. All other formulae showed higher apoptotic activities than the Free-DOX (32.5% apoptosis) (P-value >0.01, Two-way ANOVA post hoc Tukey’s test) as shown in Fig. S3.

The percentage of apoptotic cells was high in HEK 293 cells with SPG-DOX (57.3%), followed by KPG-DOX, showing 43.2% (P-value > 0.01, Two-way ANOVA post hoc Tukey’s test). The control, Free-DOX, KPG, and SPG groups exhibited low apoptotic activity of less than 20% (P-value ˂0.05, Two-way ANOVA post hoc Tukey’s test), as shown in Fig. S3.

3.7 Protein-protein interaction

Molecular docking was used to measure the binding affinity of lung cancer-derived EGFR mutants to the surface proteins of Klebsiella pneumoniae and Streptococcus pneumoniae. The wiring diagram and the 2D topology diagram of the two ligand proteins and the receptor protein obtained from the PDBsum tool are shown in Fig. 5.

Close

Fig. 5. PDBsum wiring diagram and the 2D topology diagram of the two ligand proteins: 4HWM and 3ZPP, and the receptor protein 3ITO. In the wiring diagram, the ligand proteins are in violet color, while the receptor protein was in red color. Helices are labelled H1, H2, ..., strands, represented as arrows, by their sheets A, B, β for beta turn, and ɣ for gamma turn. In the topology diagram, the red cylinder represents α-helix, the pink arrow indicates the β-strand to form β-sheet, and a small blue arrow illustrates the direction of the protein chain from N- to C-terminal.

Export to PPT

Analysis of the docking results revealed 100 poses for each interaction. The pose with the lowest binding energy, the most stable, likely binding position, and most biologically relevant was selected. The 3 D structures of the best poses in both interactions were visualized using University of California, San Francisco (UCSF) chimera 1.18 Fig. 6.

Close

Fig. 6. 3D model diagram of the best poses of the interacting proteins using University of California, San Francisco (UCSF) Chimera 1.18, violet color represents the ligand proteins, red color represents the receptor protein, and the arrows illustrate the direction of the protein chain from N- to C-terminal.

Export to PPT

The binding score (S) for the 4HWM- 2ITO interaction was -63.94 kcal/mol, while that of 3ZPP- 2ITO interaction was -61.26, which indicates very close binding affinity of both proteins to EGFR. The Gefitinib binding score was -8.18, reflecting a significantly lower affinity compared to that of the 4HWM-2ITO and 3ZPP-2ITO interactions. The 2D and 3D views of the Gefitinib interaction with EGFR protein were illustrated using Discovery studio Fig. 7.

Close

Fig. 7. The co-crystallized ligand (Gefitinib) interaction with the EGFR protein using Discovery studio. The ligand backbones are shown as gray ball and stalks and mixed with red oxygens and blue nitrogens. In each case, the conventional hydrogen and carbon-hydrogen are demonstrated as green and light green. Both alkyl and pi-alkyl bonds are shown as pink broken lines.

Export to PPT

Molecular docking analysis of protein-protein interactions was performed using a web-based PDBsum tool and database. The analysis results comparing the interface areas, the number of interface residues, salt bridges, disulfide bonds, hydrogen bonds, and non-bonded contacts are presented in Table 2.

The 4HWM-2ITO interaction exhibits a larger interface area and a higher number of salt bridges, H-bonds, and non-bonded contacts. In 4HWM- 2ITO interaction, H-bonds were formed between ASP21-TYR813, ASP21-MET987, TYR23-GLN812, and ARG111- ASP770. In the same interaction, three salt bridges were formed between ARG73-ASP830, ARG111-ASP770, and ARG111-ASP1014. In the 3ZPP-2ITO interaction, H-bonds and salt bridges were formed between GLU232-LYS823 and ASP347-LYS823. No disulfide bonds were formed in both interactions. Schematic diagrams of the two interactions are illustrated in Fig. 8.

Close

Fig. 8. Schematic diagram of interaction and the interacting residues of the ligand proteins 2OKM and 4HWM, represented as Chain A, and the 1M17 receptor protein represented as Chain B. The colored lines joining the interacting chains represent a different type of interaction, as illustrated in the above key. The amino acid residues are colored in six colors: blue, red, green, gray, violet, and orange which refer to positive (H, K, R), negative (D, E), neutral (S, T, N, Q), aliphatic (A, V, L, I, M), aromatic (F, Y, W), Pro and gly (P, G), respectively. Interactions were analyzed using PDBsum.

Export to PPT

The protein chains were represented as circles having an area proportional to the surface area of the corresponding protein chain. The interface region extent on each chain is represented by the black wedge, whose size determines the interface surface area between the interacting protein chains. The number of hydrogen bonds, salt bridges, and non-bonded contacts is also demonstrated. The interacting amino acid residues of each chain are represented by colored lines, indicating different types of interactions. The number of H-bond lines between any two residues indicates the number of potential hydrogen bonds between them. For non-bonded contacts, the width of the striped line is proportional to the number of atomic contacts.