Toxicity of amantadine hydrochloride on cultured bovine cornea endothelial cells

Amantadine hydrochloride (HCl) is commonly prescribed for treating influenza A virus infection and Parkinson’s disease. Recently, several studies have indicated that the use of amantadine HCl is associated with corneal edema; however, the cytotoxic effect of amantadine HCl has not been investigated. In the present study, the effects of amantadine HCl on cell growth, proliferation, and apoptosis in bovine cornea endothelial cells, and in vitro endothelial permeability were examined. Results showed that lower doses of amantadine HCl do not affect cell growth (≤ 20 μΜ), whereas higher doses of amantadine HCl inhibits cell growth (≥ 50 μΜ), induces apoptosis (2000 μΜ), increases sub-G1 phase growth arrest (2000 μΜ), causes DNA damage (≥ 1000 μΜ), and induces endothelial hyperpermeability (≥ 1000 μΜ) in bovine cornea endothelial cells; additionally, we also found that amantadine HCl attenuates the proliferation (≥ 200 μΜ) and arrests cell cycle at G1 phase (≥ 200 μΜ) in bovine cornea endothelial cells. In the present study, we measured the cytotoxic doses of amantadine HCl on cornea endothelial cells, which might be applied in evaluating the association of corneal edema.

www.nature.com/scientificreports/ (< 17 μM) 12 . In addition, the serum levels of amantadine HCl ranged between 2.6 (a patient who received 200 mg just for 1 day) and 16.3 μM (a patient who received 600 mg amantadine HCl for 10 days) 12 . These results indicate that amantadine HCl concentration is distributed across a wide range in different tissues. Dosage, duration of treatment, and drug-free time are all associated with mean amantadine HCl concentration. Although the mean amantadine concentration in the cornea has not yet been examined, it has been reported that amantadine has high penetrative activity into the brain after infusion in rats (brain concentration of amantadine was 16-fold higher than free concentration in serum) 13 . In recent years, an increasing number of case reports have indicated that the use of amantadine HCl is associated with corneal edema [14][15][16][17][18][19][20][21][22] . A nationwide cohort study in Taiwan also demonstrated that amantadine HCl increases the risk of corneal edema in a dose-dependent manner 11 ; however, how this occurs is still unclear. Corneal endothelium controls the water content of the corneal stroma, whereas corneal endothelial decompensation leads to overhydration of the cornea, known as corneal edema 23 . Thus, in the present study, we aimed to examine whether amantadine HCl affects cell growth, proliferation and apoptosis in bovine cornea endothelial cells.

Lower doses of amantadine HCl (≤ 20 μM) do not affect cell growth and viability in bovine cornea endothelial cells.
To examine the cytotoxicity effect of amantadine HCl on bovine cornea endothelial cells, BCE C/D-1b cells were treated with various doses of amantadine HCl (0-2000 μM) for 7 days. At 24 h, the changes in cell morphology were monitored by phase-contrast microscopy. As shown in Fig. 1A, amantadine HCl did not affect the morphology of cell growth at doses ≤ 750 μM after 24 h of treatment; however, some dead cells were found that had become detached and made clusters of a small number of cells floating in the medium when cells were treated with amantadine HCl ≥ 1000 μΜ (Fig. 1A). For cell viability, MTT assay was employed for detection from days 1 to 7. Our experimental results indicated that there was no toxicity when cells were incubated with amantadine HCl at doses ≤ 20 μΜ for 7 days (Fig. 1B). After 24 h of treatment, we found that cell viability was decreased when cells were incubated with 2000 μΜ amantadine HCl (Fig. 1B). In addition, we also found that cell viability was suppressed when cells were incubated with amantadine HCl at doses ≥ 50 μΜ for three days (Fig. 1B).

Lower doses of amantadine HCl (≤ 1000 μΜ) do not induce cell apoptosis in bovine cornea endothelial cells.
To examine whether amantadine HCl induces apoptosis in bovine cornea endothelial cells, BCE C/D-1b cells were treated with various doses of amantadine HCl (0-2000 μM) or docetaxel (DTX, 10 and 100 nM) for 24 h. Apoptotic cells were examined by Annexin V/propidium iodide (PI) staining and flow cytometry analysis. Experimental results indicated that lower doses of amantadine HCl (≤ 1000 μΜ) did not induce apoptosis, whereas higher dose of amantadine HCl (2000 μΜ) induced cell apoptosis in BCE C/D-1b cells ( Fig. 2A,B). DTX is an anti-mitotic chemotherapeutic drug that induces cell apoptosis and arrests cell cycle progression 24 ; therefore, this was used as positive control. The activity of caspase 3/7, a marker of apoptosis 25 , was also examined. BCE C/D-1b cells were treated with various doses of amantadine HCl (0-2000 μM) for 24 h. The activity of caspase 3/7 was analyzed using Caspase-Glo 3/7 assay kit. Our experimental results indicated that amantadine HCl 2000 μM significantly increased the activity of caspase 3/7, while lower doses (0-1000 μM) did not (Fig. 2C).
Lower doses of amantadine HCl (≤ 100 μM) do not affect the progression of cell cycle, but doses of amantadine HCl at 200-1000 μM induce cell cycle arrest in G1 phase in bovine cornea endothelial cells. To examine whether amantadine HCl affected the progression of cell cycle in bovine cornea endothelial cells, BCE C/D-1b cells were treated with various doses of amantadine HCl (0-2000 μM) or DTX (10 and 100 nM) for 24 h. The progression of cell cycle was measured by flow cytometry. As shown in Fig. 3, lower doses of amantadine HCl (≤ 100 μM) did not affect the progression of cell cycle; however, experimental results indicated that doses of amantadine HCl at 200-1000 μM induced G1 arrest and decreased S proportion in BCE C/D-1b cells (Fig. 3). DTX was used as positive control.
Doses of amantadine HCl at 0-750 μM do not cause DNA damage but attenuate cell proliferation of bovine cornea endothelial cells. Since doses of amantadine HCl at 200-1000 μM were found to induce G1 arrest in BCE C/D-1b cells (Fig. 3), these doses of amantadine HCl were further examined to assess whether they affected DNA integrity, DNA synthesis and cell proliferation. To test the effect of amantadine HCl on DNA damage, the alkaline comet assay was employed to detect the single-strand DNA breaks. As shown in Fig. 4, doses of amantadine HCl lower than 1000 μM (0-750 μM) had no significant effect on DNA damage visà-vis higher doses of amantadine HCl (1000-2000 μM). H 2 O 2 was used as positive control since it is a source of ROS which can cause DNA damage 26 . For the DNA synthesis, experimental results indicated that amantadine HCl significantly inhibited the EdU incorporation at doses ≥ 200 μΜ (Fig. 5A,B). In addition, the results of CFSE cell proliferation assay showed that amantadine HCl attenuated cell proliferation at doses ≥ 750 μΜ (Fig. 5C,D).
Higher doses of amantadine HCl (≥ 1000 μM) induce endothelial hyperpermeability. Endothelium maintains stromal deturgescence through barrier and pump functions. While the barrier function limits excessive fluid influx into the stroma from the anterior chamber, the fluid pump function counterbalances fluid leaks through the paracellular space 27 . Altered endothelial cell function and abnormalities or damage in the endothelial cell barrier might lead to corneal edema. To further examine whether amantadine HCl affects endothelial permeability, in vitro endothelial permeability was performed and the passage of FITC-dextran was examined. Experimental results indicated that doses of amantadine HCl ≤ 750 μM had no significant effect on www.nature.com/scientificreports/ the permeability of FITC-dextran at 1 h compared to untreated control; however, higher doses of amantadine HCl (≥ 1000 μM) significantly increased cell permeability, indicating that ≥ 1000 μM amantadine might lead to damage of the endothelial cell barrier (Fig. 6). DTX was used as positive control.

Discussion
Unlike bovine corneal endothelial cells 28 , human corneal endothelial cells do not regenerate in vivo and exhibit limited proliferative capability in vitro caused by contact-inhibited growth arrest at the G1 phase 29 . In addition, human corneal endothelial cells are not easy to obtain; therefore, in the present study, we tested the cytotoxic effect of amantadine HCl using bovine corneal endothelial cells. Our experimental results showed that acute toxicity of amantadine HCl on bovine cornea endothelial cells was not observed at doses ≤ 100 µM for 24 h; www.nature.com/scientificreports/ however, it was found that doses of amantadine HCl ≥ 200 µM induced cell cycle arrest at G1 phase and resulted in the inhibition of both DNA synthesis and cell proliferation. A recent study indicated that incubation of bovine corneal samples with 200 μM amantadine HCl for 6 h did not increase cell death compared with untreated control samples, but cell height was significantly increased compared to controls, which might result in cell death and reduce the density of corneal endothelial cells noted in patients on amantadine HCl therapy 30 . Notably, the experimental results indicated that higher doses of amantadine HCl (≥ 1000 μΜ) had significantly cytotoxicity with consistently toxic effects on bovine cornea endothelial cells including inhibiting cell growth and proliferation, inducing DNA damage and apoptosis, and increasing endothelial permeability. A previous study indicated that amantadine was transported principally across the blood-brain barrier by a saturable transport system with a one-half saturation concentration of about 1.0 mM 31 . The mean amantadine concentrations in human brain tissue ranged from 48.2 to 386 μM when the duration www.nature.com/scientificreports/ of treatment was ≥ 10 days and the drug-free time ≤ 3 days 12 . Corneal endothelium has been thought to be nonmitotic cells that have no potential in regeneration and reparation [32][33][34] . Accumulated doses of amantadine HCl might cause cornea edema. In addition, a nationwide cohort study of patients with PD in Taiwan indicated that amantadine HCl increases the risk of cornea edema in a concentration-dependent manner (a hazard ratio of 2.05 for a moderate dose (2000-4000 mg) and 2.84 for a high dose (4000 mg) 11 ; therefore, to judge the toxic effect of amantadine HCl, the exact concentration of amantadine HCl in the cornea should be further examined.
Previous studies pointed out that the CTG18.1 repeat expansion might reduce TCF4 gene expression 35 and Hessen et al. examined the copy number of CTG18.1 trinucleotide repeat in the TCF4 gene by an amantadine HCl-associated corneal edema patient 36 . Although they did not find the change on the copy number of CTG18.1 trinucleotide repeat in the TCF4 gene, genetic variation remains an important issue that might create sensitivity to amantadine HCl treatment, leading to corneal edema. In the present study, although the cytotoxic dosages of amantadine HCl were much higher than in clinical situations, accumulated doses of amantadine HCl might still pose risk to cause corneal edema in patients with rare genetic variations. www.nature.com/scientificreports/ There are several limitations in the present study. Firstly, because human cornea endothelial cells are not easy to obtain, bovine cornea endothelial cells were used to examine the toxic effect of amantadine HCl, and there might well be differences between bovine and human cornea endothelial cells. Secondly, further experiments could not proceed due to the lack of antibodies to bovine cells; thirdly, an in vitro experimental model could not achieve the cumulative dose of amantadine HCl in vivo; fourthly, the in vitro cytotoxic assays used in this study could not fully represent the real saturation of corneal edema; and finally, the duration of amantadine treatment in the present study might not represent the effects of the drug in real life due to edema formation in a physiological sense manifesting over time.
To our knowledge, this is the first study to successfully examine the cytotoxic effects of amantadine HCl using cornea endothelial cells, having performed the evaluation of these on cell growth, proliferation, apoptosis, and endothelial permeability as well as DNA integrity in bovine cornea endothelial cells.   The absolute permeability was presented as means ± SD of three independent experiments. The statistical significance was represented as follows: *p < 0.05, **p < 0.01 and ***p < 0.001 vs. untreated control. www.nature.com/scientificreports/ for 4 h at 37 °C. Afterward, the formed formazan crystals were solubilized using 100 µL hydrochloric acid-isopropanol (1 portion of 4 N HCl: 100 portion of isopropanol). After 20 min of solubilization, the absorbance of 570 nm was measured with a microplate reader (BioTek Instruments, Winooski, VT, USA).
Apoptosis assay. The apoptotic cells were detected using Annexin V and PI staining and the method was modified from a previous study 37 . Briefly, BCE C/D-1b cells were treated with different doses of amantadine HCl (0-2000 μM) for 24 h. Cells were stained with Alexa Fluor 488 Annexin V and PI in binding buffer according to the manufacturer's protocol (Thermo Fisher Scientific, Waltham, MA, USA) and analyzed by FC500 flow cytometer (Beckman-Coulter, Fullerton, CA, USA). A total of ten thousand events were collected per sample, and data were acquired and processed using CXP analysis software (Beckman-Coulter, Fullerton, CA, USA).
Caspase 3/7 activity assay. The caspase 3/7 activity assay was used to detect the activity of caspase 3 or 7 in the cells and the method was modified from a previous study 37 . Briefly, a total of 5 × 10 3 BCE C/D-1b cells were seeded in a 96-well white plate and allowed to acclimatize overnight. Afterward, cells were treated with different doses of amantadine HCl (0-2000 μM) for 24 h. Thereafter, Caspase-Glo 3/7 reagent was added to each well and gently mixed using a plate shaker at 300-500 rpm for 30 s, and then samples were incubated for 30 min at RT. Enzyme activity was directly proportional to luminescence. The luminescence intensity was detected by a luminescence microplate reader (BioTek Instruments, Winooski, VT, USA), and the data were normalized relative to the caspase 3/7 activity of cells treated with DMSO alone.
Cell cycle analysis. The cell cycle analysis was detected DNA content using flow cytometry and the method was modified from a previous study 38  In vitro permeability assay. Cell permeability assay was modified according to a previous study 40  There are several limitations in the present study. Firstly, because human cornea endothelial cells are not easy to obtain, bovine cornea endothelial cells were used to examine the toxic effect of amantadine HCl, and there might well be differences between bovine and human cornea endothelial cells. Secondly, further experiments could not proceed due to the lack of antibodies to bovine cells; thirdly, an in vitro experimental model could not achieve the cumulative dose of amantadine HCl in vivo; fourthly, the in vitro cytotoxic assays used in this www.nature.com/scientificreports/ study could not fully represent the real saturation of corneal edema; and finally, the duration of amantadine treatment in the present study might not represent the effects of the drug in real life due to edema formation in a physiological sense manifesting over time.
Statistical analysis. All data are expressed as means ± SD. Each value is the mean of three independent experiments. Statistical analysis was assessed via one-way ANOVA followed by Tukey post-hoc test using IBM SPSS Statistics v.19 (IBM Corp., Armonk, NY, USA), and the significant difference was set at *p < 0.05; **p < 0.01; ***p < 0.001.