Enhanced mitochondrial membrane potential and ATP synthesis by photobiomodulation increases viability of the auditory cell line after gentamicin-induced intrinsic apoptosis

Photobiomodulation (PBM) has been suggested to have a therapeutic effect on irreversible hearing loss induced by aminoglycosides, including gentamicin (GM). However, its intracellular mechanism(s) in GM-induced ototoxicity remain poorly understood. In the present study, we investigated the effect of PBM in GM-induced ototoxicity in auditory cells. We tried to characterize the downstream process by PBM, and the process that triggered the increased cell viability of auditory cells. As a result, the effects of PBM against GM-induced ototoxicity by increasing ATP levels and mitochondrial membrane potential was confirmed. These results suggest a theory to explain the therapeutic effects and support the use of PBM for aminoglycoside-induced hearing loss.

has been reported to have beneficial effects in medical applications, the mechanisms of PBM have not been fully understood. The prevailing theory regarding the mechanism of PBM involves bio-stimulation of mitochondria induced by laser irradiation [17][18][19] .
In the present study, we first determined the increased cell viability after PBM in GM-induced ototoxicity in auditory cells. We characterized the reverse of the downstream process of apoptosis by PBM, and then characterized the process that triggered the increase of auditory cell viability by analyzing serial levels of ATP and mitochondrial membrane potential (MMP) immediately after PBM in GM-induced ototoxicity.

Characteristics of HEI-OC1 cells and confirmation of GM-induced toxicity. To determine whether
House Ear Institute-Organ of Corti 1 (HEI-OC1) cells could serve as a model of cochlear hair cells, we conducted epifluorescence analyses using myosinVIIa staining. Par-C10 cells, a rat parotid gland cell line, were used as a negative control. HEI-OC1 cells stained for myosinVIIa, whereas Par-C10 cells did not (Fig. 1). Based on these results, we assumed that HEI-OC1 cells could be used as an in vitro model of cochlear hair cells.
To assess the possibility that HEI-OC1 cells might be damaged after GM inoculation, cell viability after GM treatment was assessed. Cell viability decreased in a dose-dependent manner. Significantly different cell viabilities among the time points were found using GM concentrations >0.1 mM (detailed statistics in Table 1). These results suggested that HEI-OC1 cells were damaged by GM and that its toxicity was dose and time dependent. Massive cell death occurred with concentrations >26.2 mM, and the largest cell viability difference among time points was observed at a GM concentration of 13.1 mM (Fig. 2). We determined that the optimal concentration for the present study would be 13.1 mM based on these results.

Confirmation of caspase activation by GM and deactivation by PBM. We investigated changes in
downstream proteins, such as cytochrome c (Cyt c), that are released by permeabilization of mitochondria, and caspases, which initiate the caspase cascade 21 . Western blotting showed that, at 6 h after laser irradiation (Fig. 4A, all analyzed protein (Cyt c, cleaved caspase 9, and cleaved caspase 3) levels increased only in GM-treated cells. However, in the GM + PBM-treated cells, the proteins were significantly reduced. Displayed blots were cropped, and full-length blots were presented in Supplementary Fig. 1. (Fig. 4B). In addition, there relative intensity differences in the Western blot analyses were observed among different groups (no GM, only GM, and GM + PBM) in all three proteins (one-way analysis of variance using the Kruskal-Wallis test: Cyt c: KW = 14.37, p = 0.0008; cleaved caspase 9: KW = 14.63, p = 0.0007; cleaved caspase 3: KW = 10.83, p = 0.0045). Using post hoc analysis, Evaluation of serial ATP levels after PBM. ATP is required for cell survival and metabolism, and GM-induced ototoxicity is known to be associated with a decrease in intracellular ATP. We therefore evaluated ATP changes after GM exposure, and also evaluated alterations in these ATP changes by PBM (Fig. 5A). After GM treatment, the ATP level showed significant decrease compared to control without GM at all time points (Mann-Whitney U-test, two-tailed, n = 8; 4 h after GM: p = 0.0021, U = 0.0; 5 h after GM: p = 0.0024, U = 0.0; 6 h after GM: p = 0.0192, U = 2.0) and showed a time-dependent increasing tendency (Fig. 5B). The increase in ATP was observed immediately and at 1 h after PBM treatment of HEI-OC1 cells with no drug treatment (Mann-Whitney U-test, two-tailed, n = 8; immediately after PBM: p = 0.0042, U = 0.0; 1 h after PBM: p = 0.0402, U = 5.5). These results corresponded to previous reports showing that GM ototoxicity was related to ATP reduction and suggested that PBM increased the ATP level in control auditory cells (Fig. 5C). In a combination of these two experiments, ATP levels were compared only between GM exposed cells and sequential GM + PBM cells, which were serially exposed. Significantly different ATP levels were observed immediately and at 1 h after PBM treatment, indicating higher ATP levels in the GM + PBM group relative to the PBM only group (Mann-Whitney U-test, two-tailed, n = 8; immediately after PBM: p = 0.0080, U = 0.0; and 1 h after PBM: p = 0.0122, U = 7.0) (Fig. 5D). Together, these results showed that ATP levels increased by PBM treatment were still effective during GM toxicity and that PBM reversed the GM-related decrease in ATP.

Reversal of PBM cell rescue by an ATP blocker.
To confirm that an increase in ATP levels 2 played an important role during the rescue process of PBM against GM toxicity, an ATP blocker was applied prior to laser irradiation (Fig. 6A). Specifically, a 4-h treatment of 0.1 μM antimycin A was applied; at this concentration and time, the potential for cell damage by ATP depletion was small ( Supplementary Fig. 2). After PBM application, compared to the group without antimycin A, the antimycin A treatment group showed significantly reduced cell viability. There was no significant difference between GM only and GM + antimycin A + PBM (one-way ANOVA p = 0.0010; F = 13.74; df = 13; n = 4-5; Tukey's Multiple Comparison Test, GM vs GM + PBM p < 0.01; GM + PBM vs GM + PBM + AA p < 0.01) (Fig. 6B). These results showed that without ATP, the rescue effect of PBM was significantly diminished, supporting the importance of ATP as a key factor for the effect of PBM.

Indirect assessment of mitochondrial membrane potential after PBM. We evaluated changes in
Rhodamine 123 by its epifluorescence intensity to monitor the membrane potential of mitochondria 22 at the a treatment when compared to controls, whereas it was significantly increased after PBM (Fig. 7B). Using these epifluorescence images, MMP was quantified and compared among groups (control, only GM, and GM + PBM) at three time points (immediately, 1 h, and 2 h after PBM). Differences among groups were observed at all time points (one-way analysis of variance with the Kruskal-Wallis test: immediately: KW = 24.76, p < 0.0001; 1 h: KW = 16.54, p = 0.0003; and 2 h: KW = 12.10, p = 0.0024). Post hoc analysis revealed that GM treatment reduced MMP at all time points (Dunn's test, p < 0.01~p < 0.001). Post hoc analysis showed a significant increase only immediately after PBM (Dunn's test, p < 0.01) (Fig. 7C). This timely PBM response of MMP led us to predict that light energy would affect MMP following further increases in ATP levels.

Discussion
Although PBM, among several phototherapeutic devices, has been approved by the US Food and Drug Administration, no theoretical background has been fully established. This study suggested a theoretical basis for the clinical application of PBM. After the report that revealed relation of nucleic factor kappaB and auditory cell damage by acoustic overexposure by Tamura et al. 23 , this is the following report to analyze how PBM affects a pathway related to inner ear ototoxicity. This information could help to facilitate the efficient use of appropriate laser techniques and therapies (e.g., molecular) to enhance its efficacy. Finally, with a theoretical background, PBM as a protective measure could potentially be used with drugs that are inevitably ototoxic in clinical settings. There are two major flows in hearing research. One is the regeneration of missing hair cells and the other is the rescue of damaged hair cells. According to previous studies, PBM seems to affect both processes of hearing recovery. Rhee, et al. suggested that PBM effect on gentamicin induced ototoxicity may be an impaction of cells www.nature.com/scientificreports www.nature.com/scientificreports/ proliferation and/or transdifferentiating 15 . One the other hand, there is another study by Rhee, et al. in the same year assessing protective effect of PBM after ototoxicity 24 . Therefore it might be necessary to clarify whether the pathway that revealed in our study is related to proliferation/transdifferentiation or protection. HEI-OC1 cell line, as explained above, is a common progenitor cell line for hair cell markers and supporting cells of the organ of Corti. It expresses hair cell markers such as Prestin, Myosin VIIa, Atoh1, BDNF, Calbindin and Calmodulin and supporting cell markers like connexin 26 and fibroblast growth factor receptor (FGF-R) 25 . Considering this potency, there is chance that the HEI-OC1 cells could have been repopulated or regenerated by laser treatment after damage rather than protected from GM. Increased cell viability and related markers after PBM without any ototoxic insults indirectly supports the regeneration theory ( Supplementary Fig. 3). On the other hand, there are findings in this study which reflected protection or rescue, as proteins that were responsible for the intrinsic apoptotic pathway were downregulated after PBM. Thus, it is not feasible to conclude at this point and we believe that both inhibition of apoptosis and proliferation/regeneration occurred after PBM and increases of ATP/MMP take an important role in this process.
Although we determined the ototoxic concentration by the actual damage model using HEI-OC1 cells, concentration of ototoxic drug applied is this study is beyond the clinical applicability. The reason for this might be related to the difference of cell structures. In organ culture or in vivo, the pathway through which GM enters the hair cell is thought be related to the Mechano-electrical transduction channel (MET channel) located the tips of stereocilia on hair cells, resulting in an abnormal rise in reactive oxygen species (ROS) and initiating cell death 26 . Unlike organ cultures or in vivo studies with MET channels, HEI-OC1 cells do not have a MET channel through which aminoglycosides can migrate. Therefore, high drug concentrations may be necessary for drugs to be toxic. Indeed, some researchers have been studied using high concentrations of drug for GM ototoxicity 27,28 .
Many published reports regarding laser therapy include references to optimal parameters, i.e., wavelength, intensity, and duration, for deriving a "good" effect; however, because PBM has frequently shown a biphasic response with different parameters, such recommendations could result in undesirable outcomes 19,29 . Laser parameters should therefore be determined carefully before actual use. In the present study, the wavelength and laser power values were determined from our previous reports and from preliminary studies. Regarding wavelength, we used 808 nm, which is widely used for protective studies of cochlea damage because of its deep penetration 15 . We used a laser power of 15 mW, which was calculated when considering the laser penetration of the inner ear (5%) 16 and the safety range of laser power in transcanal laser use (<250 mW) 30 . The duration of laser exposure was determined after evaluating the most effective exposure duration, which was 15 min (Supplementary Fig. 4).
During the process of apoptotic cell death by GM treatment, PBM exposure upregulated ATP, which is important for cell homeostasis and cell proliferation/differentiation. ATP, which is a trigger molecule used to identify a www.nature.com/scientificreports www.nature.com/scientificreports/ mechanism for PBM, is one of the major mitochondrial products. Investigating changes in mitochondrial function after PBM would therefore be necessary because MMP is a key parameter for mitochondrial function 31 . In the present study, mitochondrial function, as evaluated by mitochondrial membrane potential, was reduced after GM exposure, but returned to a normal range after PBM, suggesting that mitochondrial function loss was upregulated by PBM. Considering prior studies showing alterations of MMP after PBM 32-34 , this theory is plausible. However, detailed connections and a molecular mechanism connecting PBM and MMP have yet to be identified. Cyt c is a likely candidate protein, as one of the prevailing mechanisms of PBM is thought to be related to a Cyt c-induced increase in ATP production via effects in the mitochondrial respiratory chain 18,[35][36][37][38] . However, further molecular work is necessary to confirm this theory.
In the present study, PBM increases ATP immediately after exposure. But it reaches almost down to control level after 2 hours of exposure and there is no difference between GM and GM + PBM group at this time point. Similar results have been reported by a study showing ATP increase in the cochlear neuron after near-infrared light irradiation. According to the speculation made in this paper, the possible reason for this could be due to fast ATP consumption of cells in injury conditions 39 . Considering the fact that ATP difference between non-PBM and PBM group is dramatically reduced at 1 days after in GM treated group (not prominent in no GM treated group) in the present study as shown in Fig. 5 and supplementary Fig. 5, it is probable that reduction of ATP difference between PBM group and no-PBM group over time could be related to the insult induced ATP consumption.
In this study, we demonstrated the increased cell viability of auditory cells by PBM after GM-induced ototoxicity, which can cause permanent hair cell damage and sensorineural hearing loss, by increasing ATP levels and MMP. Moreover, we showed that PBM might be related to down regulation of apoptosis via an intrinsic caspase-dependent pathway. These results suggest a theory to support the use of PBM as a therapeutic tool for aminoglycoside-induced hearing loss. However, further studies to establish detailed connections between PBM and MMP are necessary to confirm our findings.

Materials and Methods
Cell culture. House Ear Institute-Organ of Corti 1 (HEI-OC1), a mouse auditory cell line 40   www.nature.com/scientificreports www.nature.com/scientificreports/ and GM + PBM groups. Par-C10 cells as a negative control were prepared as described in Dr. R. Biswas's report 41  The cell viability assay after GM exposure. HEI-OC1 cells were incubated in 96-well plates (SPL, Pocheon, Republic of Korea) at a density of 1 × 10 4 cells/100 μL for 24 h with fresh medium changed to medium with GM (Sigma-Aldrich, St. Louis, MO, USA, Cat# G1397). For GM exposure groups, different doses of GM solution were used starting at 50 mg/mL, followed by two-fold serial dilutions using culture medium. The control cells were treated with culture media as a vehicle that the same used to prepare gentamicin diluted solution. The cells were then incubated at 33 °C in 5% CO 2 for up to 24 h. The dose-dependent effects of GM were assessed using the 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay (Sigma-Aldrich). For the assay, 50 μL MTT solution (2 mg/mL) was added to each well. The plates were then incubated for 4 h at 33 °C under 5% CO 2 . The medium in each well was removed, and the purple formazan crystals were dissolved in 150 μL dimethyl sulfoxide (Junsei Chemical, Tokyo, Japan). After 20 s of shaking the plates in a microplate mixer (Amersham, Buckinghamshire, UK), the optical density (OD) was measured with a microplate reader (EASYUVM340; Biochrom, Cambridge, UK) at 540 nm. To assessed the cell viability, number of replicates for each experiment was performed three times with the same samples. Cell viability was calculated using the following formula: cell viability (%) = (mean OD in treated wells/mean OD in control wells) × 100. PBM protocol. A near-infrared diode laser (WonTech, Daejeon, Republic of Korea) with a wavelength (continuous wave) of 808 nm was used to irradiate the GM + laser and laser only samples. The target monolayer of cells was irradiated directly at an intensity of 15 mW for 15 min (total energy density: 13.5 J/cm 2 ) starting at 4 h after GM treatment (determined by serial testing from 0.5 h to 9 h) as stated in the Fig. 3   1 mM dithiothreitol, and 0.5 mM phenylmethanesulfonyl fluoride) containing protease and phosphatase inhibitors (1:100) (all reagents were purchased from Sigma-Aldrich) and kept on ice for 15 min. Then, 2.5% NP-40 was added, and the mixture was vigorously vortexed. The cell lysates were collected by centrifugation at 12,000 rpm for 5 min at 4 °C. The supernatants were collected and stored at −20 °C until use in the protein assay. The protein concentrations were calculated using a Pierce BCA protein assay kit (Thermo Fisher Scientific). A total of 30 μg of loaded protein was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis using a 12% separation gel. Separated proteins were electrotransferred onto polyvinylidene fluoride membranes (Bio-Rad, Hercules, CA, USA). The blots were blocked with blocking buffer containing 5% bovine serum albumin (BSA; Santa Cruz Biotechnology, Dallas, TX, USA) and 1% Tween-20 (Sigma-Aldrich) in Tris buffer (TBS) for 1 h and then incubated at 4 °C overnight with primary antibodies, following the recommended dilution range of 1:1,000-1:2,000. The following primary antibodies were from Cell Signaling Technology (Danvers, MA, US): Cyt c, caspase 9 (#9509), and caspase 3 (#9662). Membranes were washed with 0.1% TBST for 15 min, then incubated for 1 h at room temperature with horseradish peroxidase-conjugated goat anti-rabbit IgG (sc 2004; Santa Cruz Biotechnology), diluted in TBST (1:2,000), and then washed. Membranes were developed with an Enhanced Chemi Luminescence kit (GE Healthcare, Buckinghamshire, UK), and images were captured using a Kodak in vivo image analyzer (Eastman Kodak, Rochester, NY, USA) and Fuji Medical X-ray film (Fujifilm, Tokyo, Japan). (number of replicates for each experiment was three) The intensities of protein bands were assessed by ImageJ software, and histograms of relative protein expressions were prepared 42 .

Measurement of ATP levels.
To measure ATP levels, the Colorimetric ATP assay kit (Abcam, Cambridge, UK) was used according to the manufacturer's instructions. Briefly, the cells were trypsinized, centrifuged at 1,200 rpm for 5 min, and then collected at 0 h, 1 h, and 2 h after laser irradiation and resuspended in lysis buffer provided with the ATP assay kit. (number of replicates for each experiment was three) The ATP level was measured by a microplate reader using an absorbance at 540 nm. To assess the reversal of PBM cell rescue by an ATP blocker, antimycin A was dissolved it in 50 mg/mL with 95% ethanol and diluted with culture medium.

Epifluorescence analysis involving the phenotype of HEI-OC1 cells and an evaluation of MMP.
For the determination of HEI-OC1 cell characteristics, the primary antibody (rabbit myosin VIIa, 1:200; Proteus Biosciences, Ramona, CA, USA) in blocking solution composed of 1% BSA and 10% goat serum in 1 × phosphate-buffered saline (PBS, pH 7.4; Bio-Rad) was incubated for 1 h. The secondary antibody (Alexa Fluor 488-conjugated goat anti rabbit IgG, 1:200; Life Technologies, Carlsbad, CA, US) in blocking solution was used to detect the primary antibody. After incubation for 50 min and washing three times with cold 1 × PBS, the cells were mounted with Vectashield mounting medium using 4′,6-diamidino-2-phenylindole (Vector Laboratories, Burlingame, CA, USA), and then visualized using immunofluorescence microscopy (DP72; Olympus, Tokyo, Japan) at 40×.
Rhodamine 123 (Invitrogen, Waltham, MA, USA) was used to measure the membrane polarizations within mitochondria _ENREF_39. Treated cells were stained using 10 mM fluorescent dye for 15 min at 33 °C under 5% CO 2 for 0 h, 1 h, and 2 h after laser irradiation, and were then washed twice with DMEM without FBS or phenol red (Corning, Tewksbury, MA, US). The relative levels of green fluorescence in each sample were measured with a confocal microscope with emission at 488 nm under equivalent conditions (Leica510 Meta; Carl Zeiss, Oberkochen, Germany). To assess the intensity of fluorescence, the MMP intensity value within the region of interest was measured using ImageJ software ( Supplementary Fig. 6) (number of replicates for each experiment was three).
Statistical analyses. All data were analyzed using GraphPad Prism (GraphPad Software, La Jolla, CA, USA) or SPSS Statistical Software for Windows (IBM, Armonk, NY, USA). The Kolmogorov-Smirnov test was used to determine whether the data were parametric or nonparametric. Significant differences between the control and treatment groups were statistically analyzed using the t-test for parametric distributions and the Mann-Whitney U-test for nonparametric distributions, as well as analysis of variance using Tukey's multiple comparison test and two-way analysis of variance using the Bonferroni post hoc test. A value of p < 0.05 was considered statistically significant, and significance is noted as *p < 0.05; **p < 0.01; and ***p < 0.001.