The human urinary bladder cancer cell line (T24) and human urothelium cell line (SV-HUC-1) were obtained from the American Type Culture Collection (ATCC, USA). The T24 cell line was cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) Ham’s F-12 50/50 Mix (Corning, USA) and the SV-HUC-1 cell line in Kaighn’s Modification of Ham’s F-12 Medium (Corning, USA). Cells were kept at 37°C in a humidified atmosphere of 5% CO₂. Both media were supplemented with 10% fetal bovine serum (FBS), 100 U/mL of penicillin-streptomycin, and 100 U/mL of amphotericin B (Corning, USA).

Solutions ranging from 25 to 800 μg/mL were prepared by dissolving the Pharmaceutical Secondary Standard powders of moxifloxacin, ofloxacin, norfloxacin, and enrofloxacin (Sigma-Aldrich, USA) in the medium used for the appropriate cell line. Additionally, 0,05% 0,1M HCl (Reagecon, Ireland) was added to lower the pH to enable the achievement of higher solubility.

After detachment, 2.5 x 10³ T24 cells and 12.5 x 10³ SV-HUC-1 cells were seeded per well in a 96-well plate. The following day, different concentrations of moxifloxacin, ofloxacin, norfloxacin, or enrofloxacin were added (25, 50, 100, 200, 500, and 800 μg/mL), and after that, the cells were incubated for 24 and 48 hours at 37°C in 5% CO₂. Untreated cells (with the acidified medium used for dissolving the drugs) were used as a control. After the incubation, 100 μL of 1-(4,5-Dimethylthiazol-2-yl)-3,5-diphenylformazan (MTT) reagent (1 mg/mL) was added to each well, and the plate was placed in the dark for two hours at 37°C. Next, the supernatant was removed, and dimethyl sulfoxide (DMSO, POCH, Poland) was added to dissolve formazan crystals. The absorbance was measured spectrophotometrically at 570 nm and 655 nm using a Varioscan LUX plate reader (ThermoFisher Scientific, USA). Obtained absorbance results were used to calculate cell viability after fluoroquinolone treatment and to calculate the lethal concentrations (LC) causing the death of 10, 50, and 90% of cells.

To confirm the results (LC₁₀, LC₅₀, and LC₉₀) calculated based on the MTT assay, a real-time cell growth analysis was performed using the xCELLigence RTCA DP system (ACEA Bioscience, USA). This device uses data from the impedance measurement (electrical resistance) of biosensors placed at the bottom of each plate well to analyze changes in cell morphology and their proliferation rate. Analysis was performed on a 16-well E-plate (ACEA Bioscience, USA). Cells were seeded at a density of 5 x 10³ (T24) or 12 x 10³ cells per well (SV-HUC-1) and cultured in 200 μL of medium (37°C, 5% CO₂ atmosphere). After 24 hours, previously prepared LC₁₀, LC₅₀, and LC₉₀ concentrations of the appropriate antibiotics were added to the cells, and as a control, a fresh medium was used. The influence of fluoroquinolones on cell growth was analyzed for 24 and 48 hours.

Cell cycle changes were analyzed using flow cytometry and a dedicated Tali^® Cell Cycle Kit (Thermo Fisher Scientific, USA). After incubation with a fluoroquinolone, the cells were detached and centrifuged twice at 500 x g, and cells were washed with Phosphate Buffered Saline (PBS; Corning, USA) between stages. Next, 5 x 10⁵ cells were suspended in 1 mL of cold 70% ethyl alcohol (POCH, Poland) and incubated at -20°C for at least 24 hours. Then cells were again washed twice in PBS (Corning, USA) and centrifuged at 1000 x g at 4°C. The cells prepared this way were suspended in Tali^® Cell Cycle Kit solution (Thermo Fisher Scientific, USA) and incubated at room temperature in the dark for 30 minutes. Analysis was performed using the BD FACSCanto™ II flow cytometer (Becton Dickinson, United States). The obtained data were analyzed using FlowJo v10 (Becton, Dickinson, and Company, USA). The percentage of cells in the G0/G1, S, and G2/M phases was calculated using the Watson model.

The test was performed using the FITC Annexin V Apoptosis Detection Kit II (BD PharmingenTM, USA), which consists of Annexin V conjugated with fluorescein isothiocyanate (FITC) and propidium iodide (PI). After 24 hours of exposure, to calculate LC values of the fluoroquinolones, cells were detached and washed with PBS (Corning, USA). Next, the cells were centrifuged at 500 x g, and 3.5 x 10⁵ cells were resuspended in 350 μL of binding buffer. Cells were stained with 5 μL of FITC-conjugated Annexin-V (10 mg/mL) and 10 μL of PI (50 mg/mL) and then incubated for 15 minutes at room temperature in the dark. After incubation, 200 μL of binding buffer was added, and cells were analyzed using a BD FACSCanto™ II flow cytometer (Becton Dickinson, USA). Results were analyzed with FACSDiva software (Becton Dickinson, USA).

The caspase activity was measured using the Caspace-Glo^®3/7 assay (Promega, Walldorf, Germany). After seeding with the proper density on a 96-well white/clear plate (Corning, Germany), both cell lines were allowed to attach to the growth surface for 24 hours. Next, all the tested drugs’ LC concentrations calculated after 24 hours of incubation were added. Both reagents (substrate with buffer) were brought to room temperature, and after combining, an equal volume of reagent was added directly to cells in the culture medium. After that, the plate was placed in a Varioskan LUX plate reader (Thermo Fisher Scientific, USA), shaken for 30 seconds at 300 rpm, and incubated in the dark for 30 minutes, after which the luminescence signal was measured. The experiment was performed in triplicate.

For morphological analysis, T24 and SV-HUC-1 cell lines were seeded in a 12-wells plate on previously placed round glass slides at a concentration of 6 x 10³ and 15 x 10³ cells/well, respectively. The cells were treated with calculated LC concentrations of antibiotics for 24 and 48 hours. Morphological changes were examined under an inverted light microscope (DMi1, Leica, Germany).

After incubation, cells were washed with PBS and fixed with a 4% methanol-free formaldehyde solution in PBS containing 0.1% Triton-X100 (Sigma-Aldrich, USA). Next, a staining solution containing 4’,6-diamidino-2-phenyloindol (DAPI) at a concentration of 0.1 µg/mL and phalloidin-iFluor 488 (Sigma-Aldrich, USA) at a manufacturer-recommended concentration was added for 30 minutes. After that, three PBS washes were performed. Cells were visualized using an inverted fluorescence microscope (IX83, Olympus, Japan). Images were acquired in a z-stack, and deconvolution and EFI (extended focal imaging) were performed to obtain 2-dimensional images for publication.

Cells exposed to the tested fluoroquinolones at concentrations of LC₁₀, LC₅₀, and LC₉₀ were analyzed for the expression of selected genes (BAX, BCL2, TOP2A, TOP2B, and CDKN1A). Total RNA from tested cells was purified using a RNeasy Mini Kit (Qiagen, Germany). The cell pellet was disrupted with an appropriate volume of lysis buffer and processed according to the manufacturer’s protocol. The obtained material was subjected to qualitative and quantitative analysis using the NanoDrop™ Lite Spectrophotometer (Thermo Fisher Scientific, USA). Reverse transcription reactions for complementary DNA synthesis were performed using the Transcriptor First Strand cDNA Synthesis Kit with the Anchored-oligo(dT)₁₈ Primer (Roche, Switzerland) according to the manufacturer’s recommendations. Gene expression was evaluated by quantitative polymerase chain reaction (qPCR) using LightCycler^® 480 SYBR Green I Master (Roche, Switzerland) and PCR primers (Bio-Rad, USA) for selected genes with SDHA and TBP as the reference genes. The obtained gene expression results were compared to control cells and subjected to statistical analysis.

Spheroids were generated using low-attachment plates. Cells were seeded at a density of 25,000 cells per well in 96-well U-shaped PrimeSurface^® 3D plates (PHC Europe, The Netherlands), centrifuged at 200 x g for 5 minutes, and grown for 3 to 4 days. Transparent plates were used for morphology and white for luminescence analyses. CellTiter-Glo^® 3D assay (Promega, Germany) was used for the measurement of the spheroids viability. Calculated LC concentrations of tested drugs were added to the spheroids and incubated for 24 and 48 hours. The reagent was placed in a refrigerator for the night, and the next day it was prewarmed (22°C, for 30 minutes) directly before use. Fifty microliters of reagent were added to the same volume of culture medium. After that, the plate was placed in a Varioskan LUX plate reader (Thermo Fisher Scientific, Massachusetts, USA), shaken for 5 minutes at 420 rpm, and incubated in the dark for 30 minutes, after which the luminescence signal was measured. Caspase-Glo^®3/7 3D assay (Promega, Germany) was used for caspase activity measurement. To prepare the reagent for analysis, the substrate was combined with a buffer and left at room temperature before the experiment. The protocol was similar to the viability assay except for the shaking parameters (30 seconds at 500 rpm). Each experiment was performed in triplicate. The light-inverted microscope, Leica DMi1 (Leica, Germany), was used for photographic documentation. The spheroids area was measured using EPview 1.3 software (Olympus, Japan).

In the case of cytoskeleton analysis, before staining, spheroids were washed with PBS and permeabilized using 0.1% Triton-X100 for 24 hours at 4°C. F-actin was stained using Phalloidin-iFluor™ 647 (concentration following the manufacturer’s specification; Cayman Chemical, USA), and the nuclei were stained with DAPI (600 nM) for over 4 hours in 4°C. The 647 marker was used to increase sample penetration. After staining, spheroids were washed with PBS twice and transferred from their wells onto a dedicated imaging dish for confocal microscopy (µ-Dish 35 mm, high Glass Bottom; ibidi, USA). Two different imaging approaches were used to inspect the changes on a macro and micro-scale thoroughly. For the investigation of the spheroid structure, a 10x lens was used. For a more detailed examination of the cytoskeleton of the cells comprising the spheroid, a narrower field of view and higher resolution, a 63x lens, was more appropriate (LSM 900, Zeiss, Germany). For the images acquired by the 10x lens, the dedicated ZEN software deconvolution was applied. In the case of the higher magnification, airyscan software processing was performed in the workflow.

Each experiment was performed at least in triplicate. Average cell viability was expressed as a percentage relative to the control. All data were presented as means ± SD (Standard Deviation). Normality distribution was analyzed using the Shapiro-Wilk test. Statistical analysis was performed using a one-way ANOVA with Tukey post hoc (for cell viability) or two-way ANOVA with Sidak post hoc (for grouped analysis). Kruskal-Wallis with Dunn’s multiple comparison tests were used for results with non-normal distributions. Pearson comparison was performed to compare the cytotoxicity profiles of the tested drugs on all tested cell lines (GraphPad Software 8.4., USA).

The results showed that all antibiotics significantly reduce cell viability depending on dose and incubation time ( Figure 1 ). In most cases, cell viability after 48 hours of incubation was lower compared to 24 hours of incubation ( Figure 1A ). The concentration dependence on cell survival in all cases was logarithmic. Based on the obtained trend lines, a formula was generated, which enabled the calculation of the LC₁₀, LC₅₀, and LC₉₀ values ( Table 1 , Supplementary Table 1 ). Higher cytotoxicity to bladder cancer than to non-cancer urothelium cells was observed after 48 hours of incubation in the case of norfloxacin at almost all of the concentrations and enrofloxacin at the highest tested concentration. In two other tested drugs, differences between normal and cancer cell lines were observed in single concentrations ( Figure 1B ). Norfloxacin and enrofloxacin showed the strongest anticancer properties against cancer cell lines and weaker effects on normal cells. Calculated LC₉₀ values after 48-hour incubation of SV-HUC-1 were 2 times higher than for T24 cells in the case of norfloxacin, which means that non-cancer urothelium cells are more resistant to this antibiotic. Despite the lowest cytotoxic potential of all tested drugs, ofloxacin also shows a high potential due to significant differences between LC₉₀ values for T24 cells (LC_90 = 6364.7 μM) and SV-HUC-1 (LC_90 = 12386.3 μM) after 24-hour incubation with the drug.

As shown in Supplementary Figure 1 , untreated cells grew adhesively in the plate wells and were regular in shape and size. Treatment with LC₁₀ doses did not affect the morphology of both cancer and normal cells ( Supplementary Figure 1 ). A significant decrease in the number of cells and loss of contact between the cells was observed at LC₅₀ concentrations. The cells treated with LC₉₀ doses of fluoroquinolones for 24 and 48 hours lost their shape, became round, and began to detach from the well’s surface ( Supplementary Figure 1 ). Additionally, at the LC₉₀ concentration of norfloxacin and enrofloxacin crystals of these drugs started to precipitate ( Supplementary Figure 1 ).

Real-time cell growth of SV-HUC-1 and T24 lines after incubation with norfloxacin, enrofloxacin, moxifloxacin, and ofloxacin was performed using the xCELLigence RTCA system. The results confirmed that the LC values of these antibiotics, calculated based on the MTT test results, caused a decrease in cell proliferation by 10, 50, and 90% ( Supplementary Figure 2 ).

In the case of the T24 cell line, an increase in the number of cells in the G2/M phase was observed with a simultaneous decrease in the number of cells in the G1/G0 phase after exposure to all tested drugs, mainly at the LC₉₀ concentrations. However, statistically significant values were not obtained in the case of norfloxacin. This result indicates that the analyzed fluoroquinolones induce G2/M phase arrest in T24 cells. In the case of the SV-HUC-1 cell line, the effect of the tested drugs on cell cycle phase distribution was minimal. We observed an increase in the number of cells in the S phase and a simultaneous decrease in the number of cells in the G1/G0 phase after exposure to enrofloxacin. Also, a reduction in G1/G0 phases after exposure to moxifloxacin was seen. In both cases, these findings were seen after using the LC₉₀ concentration ( Table 2 , Figure 2B ).

In both tested cell lines, incubation with each of the four tested fluoroquinolones in LC₉₀ concentrations increased the number of late apoptotic cells. In the other two tested concentrations, no changes in apoptotic cell distribution were observed ( Table 2 , Figure 2A ). In the cancer cell lines, activation of caspase 3/7 was observed after incubation with a LC₅₀ of all tested drugs. In the case of the SV-HUC-1 cell line, an increase in caspase 3/7 activation was observed only in the norfloxacin LC₅₀ group ( Figure 2C ).

For all tested cell lines, the LC₉₀ of the drugs showed severe cytoskeleton degradation and a decrease in the number of cells visible in the field of view compared to the control. In most cases, it was impossible to observe F-actin stress fibers due to severe degradation, with norfloxacin in the SV-HUC1 cell line being the exception. LC₉₀ of ofloxacin and moxifloxacin was also responsible for severe morphological cell nucleus changes ( Figure 3 ; Supplementary Figure 3 ).

Both LC₉₀ and LC₅₀ of all tested drugs decreased the staining efficiency of phalloidin. Staining uniformity was hindered, and the resulting signal was weaker compared to controls (data not shown). LC₁₀ of ofloxacin and norfloxacin had the least effect on the morphological changes of the SV-HUC1 cell line, while norfloxacin had the least impact on T24 cells. Moxifloxacin at all tested concentrations caused the detachment of the remaining T24 cells and hindered the fixation process. Similar observations were made for T24 cells exposed to LC₉₀ of ofloxacin ( Figure 3 ; Supplementary Figure 3 ).

The cytotoxic profile analysis of all tested fluoroquinolones showed similar effects on both tested cell lines. The most significant differences were observed in the case of ofloxacin, which showed the weakest effect, especially against the T24 cell line, after 48 hours of incubation. Moxifloxacin also showed a weaker effect, however, its cytotoxic profile in higher doses was more comparable to norfloxacin and enrofloxacin than ofloxacin. In the case of norfloxacin and enrofloxacin, differences were rarely observed and mostly with low doses of drugs. However, after the 48-hour incubation, enrofloxacin was more effective in reducing non-cancer cell line viability ( Figure 4A ). These observed results were confirmed by Pearson comparisons of the cytotoxic profiles. Correlation between all tested drugs was very high, almost all combinations reached values above 0.9. This analysis showed a very high similarity of cytotoxic profiles between the tested drugs in both incubation times. The lowest value, below 0.9, was observed in the case of ofloxacin, which confirmed previous results indicating the insufficient effectiveness of this drug ( Figure 4B ).

Overexpression of the BAX gene was observed in the highest tested concentration of moxifloxacin and ofloxacin in both cell lines. This effect was evident in the case of moxifloxacin in the T24 cell line. The BCL-2 gene was overexpressed in non-cancerous cell lines after treatment with a LC₉₀. In the case of cancer cells, only a slight increase was observed in the LC₉₀ ofloxacin group. Tested drugs reduced the expression of TOP2A and TOP2B genes at the highest concentration. This effect was not observed in the case of ofloxacin in the T24 cell line and enrofloxacin for the TOP2A gene in the SV-HUC-1 cell line. In the CDKN1 gene, we observed an increased expression with the LC₅₀, mainly in cancer cells, and a decrease in its expression in LC₉₀ (mainly in non-cancer cells). In the SV-HUC-1 cell line treated with LC₉₀ of norfloxacin and moxifloxacin, the product of this gene was not detected ( Figure 5 ).

The cytotoxic effect of fluoroquinolones was more pronounced in cancer cell lines in 3D cultures. This effect was visible, especially with the LC₉₀ in T24 cell lines after both incubation times, except for the LC₉₀ of enrofloxacin after a 24-hour incubation, in which cell viability was higher ( Figure 6 ). Together decreased viability and increased spheroid diameter and caspase 3/7 activity was observed ( Figures 6 – 8 ). LC₅₀ did not reduce cell viability by 50% in 3D cultures. In the case of the SV-HUC-1 cell line, the cytotoxic effect was visible only in the LC₉₀ moxifloxacin group after 24 hours and LC₉₀ ofloxacin after 48 hours ( Figure 6 ).

LC₉₀ of all tested compounds affected the structural integrity of the spheroids grown from T24 and SV-HUC1 cell lines. LC₉₀ of moxifloxacin had the most severe effect on the cytoskeleton of both cell lines in the 3D spheroid model. Staining revealed a complex degradation of stress fibers at LC₅₀ regardless of exposure time. DAPI staining showed a noticeable decrease in nucleus complexity, causing degradation and multiple deformations compared to the control. LC₉₀ of enrofloxacin and ofloxacin substantially affected the F-actin cytoskeleton of T24 and SV-HUC-1 spheroids after 24 hours of incubation. Minor differences between these two drugs were visible after 48 hours of incubation, where enrofloxacin had a more significant effect. LC₅₀ of enrofloxacin and ofloxacin also caused visible cytoskeleton degradation. This concentration had the most considerable impact on T24 spheroids after 48 hours of incubation. LC₅₀ of both drugs affected the nuclei morphology of SV-HUC-1 spheroids strongly after 24 and 48 hours. This effect was also visible for T24 spheroids, but a more decisive response was observed after 48 hours. Norfloxacin caused similar changes in both tested cell lines regardless of exposure. In the case of this drug, we observed more severe nuclei morphology changes than with the other tested drugs. It was evident when imaging SV-HUC-1 cells, where the LC₁₀ caused noticeable artifacts of the cell nuclei. On the contrary, norfloxacin did not cause F-actin cytoskeleton changes even at LC₅₀ for SV-HUC-1, but we did observe those changes for T24 cell spheroids. The morphological cytoskeleton changes were very pronounced for LC₅₀ and LC₉₀. Exposure to LC₅₀ caused almost no changes in those parameters, whereas LC₉₀ caused significant changes ( Figures 9 , 10 ; Supplementary Figures 4 - 7 ).