Paeoniflorin protects spiral ganglion neurons from cisplatin-induced ototoxicity: Possible relation to PINK1/BAD pathway

2.1 Animals and treatments

We purchased the C57BL/6 mice from the Animal Center of Shandong University (Jinan, China) and conducted all animal experiments according to protocols approved by the Animal Care Committee of Shandong University following the guidance for the care and use of Laboratory Animal for Research Purposes. Experiments in vivo studies were implemented on age and sex-matched postnatal day (P) 30 C57BL/6 mice. 3 mg/kg cisplatin for consecutive 7 days has been widely used in setting up animal models with cisplatin-induced ototoxicity.27, 28 It has been reported that intraperitoneal injection with paeoniflorin (15, 30 mg/kg) could suppress cell apoptosis and the anti-apoptotic effect of PF was better in 30 mg/kg than that in 15 mg/kg.29 Therefore, in this work, P30 mice in the control group received normal saline and mice in PF group were given 30 mg/kg PF ip for consecutive 7 days. Mice in cisplatin group received cisplatin injection (3 mg/kg ip) daily for 7 days and mice in PF plus cisplatin group were administered 30 mg/kg PF ip for 2 hours before the daily injection of cisplatin.

2.2 Organotypic culture of neonatal mouse cochleae and drug treatment

The C57BL/6 mice were decapitated at P3 and collected only the middle cochlear turn to keep the consistency. Dissection steps and cochlear middle turn culture protocols were conducted as mentioned in previous studies.27, 30, 31 Briefly, we removed the cochlear capsules from the temporal bones to expose the membranous labyrinth, wiped out the stria vascularis and placed the middle turn cochlear explants containing the SGNs onto 10 mm coverslips. According to the previous studies from our laboratory, 50 μmol/L cisplatin treatment for 48 hours induced the most degenerative changes in SGNs compared to the 25 and 150 μmol/L.27 Another study has observed that PF at 3, 10 or 30 μmol/L in the media significantly suppresses cell apoptosis and concentrations exceeding 30 μmol/L have no further cytoprotective effect.32 Therefore, in our experiments, the middle turn explants in control group received no treatment and explants in PF group were given 30 μmol/L PF for 50 hours. We treated the explants in cisplatin group with 50 μmol/L cisplatin for 48 hours (Sigma). For PF plus cisplatin group, we pre-treated the explants with 30 μmol/L PF for 2 hours 33 followed by PF plus cisplatin for 48 hours. In vitro study, the proteins for Western blot were extracted from the middle turns, from which the hair cells parts were removed as much as possible.

2.3 Tissue preparation for frozen sections

The harvested cochleae were fixed at 4°C overnight with 4% paraformaldehyde and decalcified in 10% EDTA. Then we incubated these samples in 10%, 20% and 30% sucrose and finally the OCT compound (Tissue-Tek, Sakura Finetek). Subsequently, we adjusted these samples into right position, frozen them on dry ice and cut them into 7 μm sections.

2.4 Protein extraction and Western blot analysis

The tissues were collected quickly and lysed with RIPA (Protein Biotechnology, China), which contained the protease inhibitor cocktail (Sigma). Then, we used the BCA Protein Assay Kit (Protein Biotechnology) to estimate the protein concentration. Lysates were separated on 10% SDS-PAGE gels and transferred to PVDF membrane, which were blocked with milk and incubated with primary antibodies β-actin (1:2000, ZSGB-BIO, China), Bax (1:1000, Cell Signaling Technology), cleaved caspase-3 (1:1000, Cell Signaling Technology), Bad (1:1000, Abcam) and PINK1 (1:1000, Abcam) overnight at 4°C. The immunoblots were detected by the ECL kit (Santa Cruz) and Image Laboratory analysis software (Bio-Rad).

2.5 Immunofluorescence staining

The procedures were performed as previously mentioned33 and the samples were incubated with the primary antibodies mentioned blow: anti-NeuN (1:1000, Cell Signaling Technology), anti-Bad (1:500, Abcam) and anti-Tuj 1 (1:1000, Neuromics). Then, on the following day, we treated the samples with secondary antibody (1:1000, Invitrogen) along with DAPI (1:800, Invitrogen).

To detect mitochondrial ROS levels, we used Mito-SOX Red staining (Life Technologies) on the basis of the manufacturer's instructions. And we took advantage of the TUNEL assay (Invitrogen) to detect apoptosis.

2.6 Spiral ganglion neuron counting

We used Image J software to quantify the immunostaining positive cells. In vivo study, we measured the relative length of the scale bar (20 μm) in the figures and set scale to 20 μm in the Image J software. Then, we calculated the triangular area where SGN localized and used the ‘Point selections’ tool in the Image J to count the immunostaining positive cells. The SGN density per unit area = total immunostaining positive cell number/total triangular area of SGN. In vitro study, SGNs in which the nucleus comprised 40% of the soma area were counted in each optical section using the Image J software. The total number of SGNs in each spiral ganglion explant was obtained by summing the SGN counts in all consecutive sections, and the SGN density per unit area was calculated as mentioned before.

2.7 Statistical analyses

The data we shown were the summary from at least three individual experiments and were presented as the mean ± SD. We performed a one-way ANOVA followed by a LSD's multiple comparison test when comparing more than two groups. And statistical significance was determined at P < 0.05.

3.1 PF ameliorates cisplatin-elicited SGN loss in vivo

After treatment for 7 consecutive days, the ABR thresholds in cisplatin group were significantly increased than those in control group as shown in Figure S1. And then the cochleae were dissected out and the middle-turn cochlear sections were selected as the representative. We used Tuj 1 as the SGNs marker and found that there was no significant difference between the control group and the PF group with regard to the morphology and survival number of SGNs. Immunofluorescence results showed that there was obvious SGN loss in the cisplatin group in contrast to the high-density SGNs in the control group, while SGNs with PF pre-treatment exhibited less loss than those in cisplatin group (Figure 1A). Then we calculated the Tuj 1-positive cells in middle-turn cochlear sections and detected that in cisplatin group the mean density of SGNs was decreased contrasted to the control group (##P < 0.01, n = 3). In PF pre-treated mice, the SGN survival was decreased compared to that in the control group but increased than the group submitted to cisplatin only (#P < 0.05, *P < 0.05, n = 3) (Figure 1B).

3.2 PF alleviates the cisplatin-elicited SGN injury in vitro

According to the previous study, we treated the middle turn explants with the 50 μmol/L cisplatin for 48 hours. The somata of SGNs in control group and PF group were large and round or oval-shaped with intense Tuj 1 cytoplasm labelling and weak nucleus labelling. The peripheral auditory nerve fibres (ANFs), which protrude out radially from SGNs to HCs, were also labelled with Tuj 1 and organized into smooth and thick fascicles. In the cisplatin group, the number of SGNs were obviously decreased, the somata of the remaining SGNs were shrunken and shrivelled and the ANFs were fragmented and in disordered arrangement. However, in PF plus cisplatin group, the SGNs showed less loss and the somata were larger compared to the group only received cisplatin (Figure 2A). Subsequently, we counted the SGNs survival and found that the SGN survival in cisplatin group was remarkably reduced (###P < 0.001, n = 4) while the PF pre-treated group showed less SGN loss contrasted to the group only received cisplatin (*P < 0.05, n = 4) (Figure 2 B).

3.3 PF extenuates the cisplatin-induced SGN apoptosis in vivo

We used TUNEL assay to detect SGN apoptosis in middle-turn cochlear sections and used NeuN to label SGNs. In cisplatin group, strongly TUNEL positive staining could be found in SGNs and supporting cells, while PF plus cisplatin group showed less positive staining (Figure 3A). The TUNEL/NeuN double-positive and TUNEL-positive cells were dramatically enhanced in explants subjected to cisplatin only (##P < 0.01, ###P < 0.001, n = 4), whereas these numbers in PF pre-treated mice were remarkably reduced contrasted to the group only received cisplatin (*P < 0.05, **P < 0.01, n = 4) (Figure 3B).

3.4 PF protects SGNs against the cisplatin-induced apoptosis in vitro

We used TUNEL assay to measure the cisplatin-induced apoptosis in SGNs. SGNs that were double labelled by Tuj 1 and TUNEL were considered to be apoptotic SGNs and there was no double-positive SGN in control group. We found the TUNEL/Tuj 1 double-positive cell percentage was raised in the explants only received cisplatin (##P < 0.01, n = 4) while this percentage was reduced contrasted to the group subjected to cisplatin only (*P < 0.05, n = 4) (Figure 4A,B).

3.5 PF inhibits the mitochondrial apoptotic pathway both in vivo and in vitro

Given that the aforementioned results suggesting apoptosis is activated in cisplatin-induced SGN damage as well as PF can ameliorate cisplatin-elicited apoptosis, we subsequently explored the mechanism underlying the protective effect of PF in cisplatin ototoxicity. Western blot showed that in cisplatin group Bax and cleaved caspase-3 were remarkably raised in vivo and in vitro (#P < 0.05, ##P < 0.01, ###P < 0.001, n = 3), whereas in PF pre-treatment group Bax and cleaved caspase-3 were reduced contrasted to the group only subjected to cisplatin (*P < 0.05) (Figure 5A,B).

3.6 PF mitigates the BAD activation in SGNs after cisplatin exposure

Next, we investigated the expression of BAD through immunofluorescence and Western blot. Immunofluorescence results showed that in control group the fluorescence of BAD was very weak, while in cisplatin group the expression of BAD was strongly positive and distributed in a spotty manner around the nuclei, where typically mitochondria are clustered in the cell. In PF plus cisplatin group, the immunostaining of BAD was relatively weaker than that in cisplatin group (Figure 6A). We also found that in cisplatin group the BAD/Tuj 1 double-positive cell percentage was remarkably raised (##P < 0.01, n = 4) while in PF pre-treatment group the percentage was reduced than that in the cisplatin group (*P < 0.05, n = 4) (Figure 6A,B). Moreover, BAD protein level was dramatically enhanced in cisplatin group contrasted to the control group (###P < 0.001, n = 4), while the BAD expression in PF plus cisplatin group was reduced than the group subjected to cisplatin only (*P < 0.05, n = 4) (Figure 6C).

3.7 PINK1 is relatively increased in PF pre-treated SGNs after cisplatin insult

Analysis of the protein level of PINK1 in different groups showed that the PINK1 expression was remarkably decreased in cisplatin group (##P < 0.01, n = 4), while the PINK1 expression was increased in PF plus cisplatin in comparison to the group subjected to cisplatin only (*P < 0.05, n = 4) (Figure 7A).

3.8 PF pre-treatment reduces mitochondrial ROS levels in SGNs after cisplatin insult

Lastly, Mito-SOX red staining was applied to estimate the ROS level of SGNs, and we found that in cisplatin group the immunofluorescence was more apparent in contrast to the PF pre-treatment group. We also calculated the Mito-SOX Red-positive SGN percentage in each group and found that in cisplatin group the percentage was increased (###P < 0.001, n = 4) (Figure 7C), whereas this percentage was decreased in PF pre-treatment group than the group subjected to cisplatin only (**P < 0.01, n = 4) (Figure 7C).