Thermosensitive hydrogel containing ethosuximide-loaded multivesicular liposomes attenuates age-related hearing loss in C57BL/6J mice

Ethosuximide is the first drug reported to protect against age-related hearing loss, but its benefits are hampered by the pronounced side effects generated through systemic administration. We prepared a thermosensitive hydrogel containing ethosuximide-encapsulated multivesicular liposomes (ethosuximide-loaded MVLs-Gel) and evaluated its functional and histological effects on age-related hearing loss in C57BL/6J mice. The MVLs-Gel showed slow sustained-release characteristics up to over 120 h. After 8 weeks of treatment, compared to the oral systemic administration of ethosuximide, intratympanic ethosuximide-loaded MVLs-Gel injection dramatically reduced the loss of age-related spiral ganglion neurons in the apical turns of the mice (low-frequency regions, p < 0.05). Correspondingly, compared to the oral systemic administration group, the intratympanic ethosuximide-loaded MVLs-Gel injection group showed significantly lower auditory brainstem response threshold shifts at stimulus frequencies of 4, 8, and 16 kHz (low-and middle-frequency regions, p < 0.05). In conclusion, intratympanic ethosuximide-loaded MVLs-Gel injection can reach the apical turn of the cochlea, which is extremely difficult with oral systemic administration of the drug. The ethosuximide-loaded MVLs-Gel, as a novel intratympanic sustained-release drug delivery system, attenuated age-related hearing loss in C57BL/6J mice.


Introduction
Age-related hearing loss (ARHL) is the most prevalent sensory disorder in elderly individuals [1].It affects quality of life and mental health and may result in social isolation, depression, or cognitive disorders [2].Its estimated prevalence is as high as 60 % for people aged 71-80 years [3].ARHL is a complex disease caused by aging and environmental, genetic, and other factors and is closely related to oxidative stress, gene mutation, cochlear microcirculation, ion channels, hormones, immunity, diet, and so on [4].ARHL can be caused by various reasons, but the underlying mechanisms remain to be well understood [5].Currently, no medical treatment is available for this age-related sensory dysfunction.Research on the prevention or treatment of presbycusis has mainly focused on oxidative stress, calcium channels, vasodilators, and so on.Sirt3 (silent mating-type information regulation 2 homolog 3) has been shown to enhance the mitochondrial glutathione antioxidant defense system during caloric restriction and to prevent ARHL under caloric restriction [6].
Calcium imbalance plays an important role in age-related neurodegenerative diseases [7][8][9].Voltage-gated calcium (Cav) channels include three major families: Cav1, Cav2, and Cav3.The Cav3 channel, which belongs to low-voltage activated calcium channels, plays a significant role in regulating neuronal excitability.In their study, Sang et al. found that the activation of this channel was responsible for ARHL in NOD/LtJ mice [10].Geng et al. found that Cav3 was upregulated in spiral ganglion neurons (SGNs), which may lead to the degeneration of SGNs in ARHL mice [11].Lei et al. indicated that the oral systemic administration of the Cav3 channel blocker ethosuximide for 8 weeks could delay age-related loss of SGNs [12].The Cav3 channel comprises three subtypes: Cav3.1, Cav3.2, and Cav3.3.In the cochlea of C57BL/6J mice, Cav3.1 is expressed in the organ of Corti, the lateral wall, and SGNs, and this expression significantly decreases in the organ of Corti and SGNs of aged mice.This indicates that Cav3.1 plays a protective role in ARHL [13].Cav3.2 is highly expressed in SGNs and is significantly higher in older mice, which indicates that it plays a destructive role in ARHL.As for Cav3.3, it does not change with age [11].Therefore, the Cav3 channel blocker ethosuximide can significantly attenuate ARHL, most likely through its effects on the Cav3.2 subunit of SGNs.Therefore, when evaluating the otoprotective effects of ethosuximide in C57BL/6J mice with ARHL, we focused on SGNs.
The Cav3 channel blocker ethosuximide, which is a clinically approved antiepileptic drug, is the first drug reported to protect against ARHL.However, long-term systemic drug administration of ethosuximide may cause serious side effects, including leukopenia, aplastic anemia, and liver or kidney damage.Systemic therapy is also characterized by limited drug delivery to the cochlea because of the minimal blood flow and the blood-perilymph barrier [14,15].Currently, intratympanic drug delivery to the inner ear is increasingly applied in both clinical and animal studies.However, drugs delivered to the inner ear could be lost through the eustachian tube or cleared by the middle ear mucosa [16].Recently, many thermosensitive and sustained release systems have been developed to improve inner ear drug delivery.Most of these studies treated mice at the age of 12-32 weeks for 8-12 weeks [12,17,18].
FDA-approved poloxamer 407 (P407) exhibits thermos-reversible and mucoadhesive properties, which are important for easy and efficient intratympanic drug delivery [19].The aqueous poloxamer turns to micellar gel when the temperature exceeds the solution-gel transition temperature, thus extending the residence time in the tympanic cavity and reducing clearance from the eustachian tube [20].Prolonged residence time ensures sustained drug release into the cochlea.Furthermore, the administration of P407 is considered safe in the cochlea [21,22].
Our recent research showed that the P407 hydrogel containing dexamethasone-loaded multivesicular liposomes (MVLs) exhibited high stability and safety as a novel intratympanic sustained-release drug delivery system [25].MVLs have many separate aqueous compartments divided by nonconcentric lipid bilayers.Numerous internal aqueous chambers can provide high encapsulation efficiency for water-soluble drugs, and a nonconcentric network can confer a longer duration of drug release [23].The particle size of MVLs is approximately 1-100 μm.
A larger particle size can prevent quick cleanup by mucosa macrophages and enable sustained drug release [24].
In this study, combining P407 and MVLs, we prepared a thermosensitive hydrogel containing ethosuximide-encapsulated MVLs (ethosuximide-loaded MVLs-Gel) and investigated its drug entrapment efficiency and release properties.We further evaluated age-related loss of hearing and SGN degeneration in C57BL/6J mice following intratympanic injection of C57BL/6J with ethosuximide-loaded MVLs-Gel.To the best of our knowledge, this is the first report of MVL use in intratympanic drug delivery for ARHL therapy.

Preparation of ethosuximide-loaded MVLs
Ethosuximide and 1,2-dipalmitoyl-sn-3-phosphoglycerol (DPPG) were ordered from Sigma-Aldrich.Sucrose, glucose, L-arginine, L-lysine, gluconate, soybean lecithin, glycerol trioleate, cholesterol, dichloromethane, and diethyl ether were purchased from Sinpharm Chemical Reagent (Shanghai, China).MVLs were prepared according to the procedures described in our previous study [25].In brief, an inner aqueous solution was prepared by adding ethosuximide (2 g), sucrose (1 g), and L-arginine (140 mg) to 20 mL of water.We prepared the lipid solution with 240 mg lecithin, 80 mg glycerol trioleate, 120 mg cholesterol, and 40 mg DPPG in 20 mL of dichloromethane and diethyl ether (1:1).The lipid solution was emulsified with the inner aqueous solution to create a water-in-oil emulsion (the first emulsion).The solution was further emulsified for 2 min using an ultrasonic homogenizer (NingBo Scientz Biotechnology Co., Ltd.).This initial emulsion was then blended with an outer aqueous solution of 5 % glucose containing 365 mg lysine in mL of water at 3000 rpm for 0.5 min.To produce the water-oil-water double emulsion (the second emulsion), the organic solvent was eliminated using a vacuum rotary evaporator at 60 rpm for 1 h in 40℃ water.The resulting ethosuximide-loaded MVLs were collected following centrifugation at 600 × g.

Encapsulation efficiency determination
A standard curve was plotted from certain concentrations of ethosuximide and used to determine the ethosuximide concentration in MVLs.The MVL size was measured using a Mastersizer 3000 (Malvern, UK).The MVLs were centrifuged at 600 × g for 10 min to separate the free ethosuximide.The solution was treated with 2 mL methanol, sonicated for 10 min, and then measured using high-performance liquid chromatography (HPLC).

Ethosuximide-loaded MVLs-Gel preparation
Ethosuximide-loaded MVLs-Gel was prepared using the cold method described in our previous study [25].We chose low-speed centrifugation (600 g) to isolate ethosuximide-loaded MVLs from a dissociative ethosuximide solution.We added phosphate-buffered solution (PBS) to the ethosuximide-loaded MVLs to produce the ethosuximide-loaded MVL suspension.A distinct proportion of P407 and poloxamer 188 was put into the ethosuximide-loaded MVL suspension.We calculated the gelation temperature using a stirring method.The samples were kept under refrigeration before administration.

In vitro drug release study
Conventional Franz diffusion cells cannot reflect the relative size of the round window membrane and tympanic cavity.Thus, we chose the modified diffusion cells reported in [26].For the modified diffusion cells, we used a silica gel membrane (thickness of 0.5 mm) with a hole of 0.3 mm 2 in the center and a cellulose membrane (thickness of 0.22 μm) to isolate the receptor and donor cells.For the in vitro release experiments, 1 mL of the prepared solution was added to the donor cell, and 2.5 mL of PBS containing 0.2 % sodium azide was applied to the receptor.We sealed the donor cell with parafilm and immersed it in 37 • C water with continuous magnetic rotating.For drug measurement using HPLC, we collected 0.1 mL samples, centrifuged them at 600 × g for min, and put the same-size receptor solution back to the receptor cell.We then washed the donor cells with PBS and collected the solution to quantify the residual ethosuximide.Each sample was measured in triplicate.

Animal experiment group
All animal procedures were approved by the Institutional Review Board of the Eye, Ear, Nose, and Throat Hospital at Fudan University (Shanghai, China).We purchased 4-week-old C57BL/6J mice from Shanghai SLAC Laboratory Animal Co. Ltd. (Shanghai, China).After being fed for 8 weeks, these 12-week-old mice were randomly assigned to three groups: (1) mice treated with an intratympanic injection of normal saline (n = 10), (2) mice treated with an intratympanic injection of ethosuximide-loaded MVLs-Gel (n = 10), and (3) mice treated with oral systemic administration of ethosuximide (n = 10).For oral systemic administration, each mouse received ethosuximide (200 mg/kg weight per day) in drinking water.The drug solution was adjusted to body weight, maintained in black bottles, and changed once every three days.The ethosuximide was kept in black bottles and replaced every three days.
W. Li et al.

Intratympanic injection procedure
We anesthetized the mice by intraperitoneal injection of Zoletil (0.6 mL/kg).The tympanic membrane was exposed using a surgical microscope.We perforated the left tympanic membrane anterior to the malleus with a sterile 30-gauge needle and injected the solution through the posterosuperior quadrant of the tympanic membrane until the tympanic cavity was nearly full.We kept the left ear upward for 30 min.Intratympanic injections were repeated once every two weeks for a total of three times (Fig. 1).

Evaluation of hearing function
We performed hearing tests before and after the experiment.The auditory brainstem response (ABR) was recorded for hearing assessment, as described in our previous study [27].We anesthetized the mice with ketamine (100 mg/kg) and xylazine (25 mg/kg) through intramuscular administration.Tone burst stimuli (5 ms duration, 0.5 ms rise-fall time) were generated, and each average response was based on 1000 repetitive stimuli applied at frequencies of 4, 8, 16, 24, and 32 kHz.Click stimulus (0.1 ms duration, 10/s rate) was reduced in intensity from 100 dB sound pressure level in 5 dB steps.

Quantification of SGNs
When the mice were 20 weeks old, immediately after the final ABR was completed, each mouse was euthanized with ketamine.The cochleae were extracted and immersed in 4 % phosphate-buffered paraformaldehyde (pH 7.4).Microscopically, we opened the round window and oval window, and then punctured a small hole at the apex of the cochlea.We perfused the cochlea with 4 % paraformaldehyde in 0.1 M PBS through the apical hole, round window, and oval window.After overnight fixation, decalcification of the cochleae was performed with 16.8 % EDTA in PBS (pH = 7.4) for 2 days at 4 • C.
The cryosection was prepared based on a previous study [28].The samples were immersed in 15 % sucrose for 2 h and then 30 % sucrose in PBS at 4℃ overnight, and then embedded in an optimal cutting temperature (OCT) compound (Sakura Finetek) overnight at 4℃.Serial frozen sections of 10 μm thickness were prepared for immunofluorescence.Every fourth section was mounted on a glass slide and stained with toluidine blue.Six of the most mid-modiolar sections were randomly selected for each animal and used for the quantitative analysis of the SGNs.All SGNs with clear nuclei and cytoplasm in Rosenthal's canal of the base, middle, and apex were counted.The area of Rosenthal's canal was measured using ImageJ.The number of SGNs in each turn was calculated per mm 2 .

Statistical analysis
The data were described as the mean and standard deviation (SD).A repeated measures analysis of variance (ANOVA) test was employed to compare the differences.The level of statistical significance was defined as α = 0.05 of two-sided probability.All statistical analyses were performed using the Prism 8 statistical analysis program (GraphPad San Diego, CA, USA).

Preparation of ethosuximide-loaded MVLs-Gel
The formulation of ethosuximide-loaded MVLs was optimized using a single factor and the orthogonal experimental method.The morphology of the ethosuximide-loaded MVLs is shown in Fig. 2. The mean diameter of the MVLs was 14.7 μm, and more than 95 % of the particles were between 10 μm and 90 μm.The encapsulation efficiency of the MVLs was 61.2 % ± 2.9 %.
According to our tests, a 20 % (w/w) stock solution of P407 was liquid when refrigerated but solid at 21℃.A 26 % (w/w) P407 solution became a gel at 16℃.Poloxamer 188 (P188), a material with higher polyethylene oxide, is commonly used to change the gelation temperature of P407.Thus, P188 was used to obtain an appropriate gelling temperature.By adjusting the concentration of P407 and its ratio to that of P188, the gelling temperature was changed from 32.2℃ to 55.4℃ (Table 1).When the ratio was 18:10:72, the gelling temperature reached 35.5℃.This is appropriate because 35.5 • C is close to the normal ear temperature.

Drug concentration in the round window membrane model
As shown in Fig. 3, the ethosuximide-loaded MVLs-Gel group showed the slowest sustained release, up to over 120 h, with 73.3 % of ethosuximide released from the formulations.However, ethosuximide MVLs showed a sustained release of 70 % after 48 h, whereas the ethosuximide hydrogel encapsulating P407/P188 (18 %/10 %) showed a sustained release of 75.8 % after 10 h.

Delay in age-related increases in ABR thresholds
We next investigated whether the local delivery of ethosuximideloaded MVLs-Gel or the oral systemic administration of ethosuximide can protect against age-related increases in ABR thresholds.The baseline ABR thresholds (12 weeks old) were not significantly different.Mice were injected once every two weeks, and three injections were given in total.Two weeks after three intratympanic injections, we conducted hearing tests again (20 weeks old, n = 10/group).As shown in Fig. 4, after oral systemic ethosuximide administration (n = 10/group) to 12week-old mice for 8 weeks, the ABR threshold shifts were significantly lower than the thresholds in the intratympanic normal saline injection group (n = 10/group) at stimulus frequencies of 8, 16, and 32 kHz (p <  For the animals in the intratympanic ethosuximide-loaded MVLs-Gel injection group, significantly less ABR threshold shifts were observed compared to the animals in the intratympanic normal saline injection group at all stimulus frequencies of 4, 8, 16, and 32 kHz (p < 0.05).Compared to the ABR threshold shifts in the oral systemic ethosuximidetreated group, the intratympanic ethosuximide-loaded MVLs-Gel injection group showed significantly lower ABR threshold shifts at stimulus frequencies of 4, 8, and 16 kHz (p < 0.05); however, there was no significant difference at 32 kHz (p > 0.05) between the oral systemic and intratympanic MVLs-Gel injection groups.This indicates that ethosuximide can reach the apical turn of the cochlea through intratympanic sustained-release ethosuximide-loaded MVLs-Gel injection.

Prevention of age-related increases in SGN loss
To investigate the influence of ethosuximide-loaded MVLs-Gel on SGN survival, we examined the density and qualitative morphology of SGNs in mid-modiolar sections.As shown in Fig. 5, the SGN density of basal and middle turns in the oral systemic ethosuximide administration group (n = 10/group) significantly increased compared to the intratympanic saline solution injection group (n = 10/group) (p < 0.01), although there was no significant difference in the apical turn (p = 0.499).However, compared to the intratympanic normal saline injection group, SGN density in the intratympanic ethosuximide-loaded MVLs-Gel injection group demonstrated significantly less SGN degeneration in the basal, middle, and apical turns (p < 0.01).Age-related SGN loss was significantly reduced in the apical turns of mice treated with intratympanic ethosuximide-loaded MVLs-Gel injection compared to oral systemic ethosuximide administration (p < 0.05).

Discussion
The clinical manifestations of ARHL are progressive bilateral symmetrical hearing loss, and the hearing curve is mostly slope-shaped with high-frequency reduction.In a classic study of human hearing in populations living in isolated nonindustrialized societies, progressive hightone hearing loss characterizing ARHL was not seen.Kujawa and Liberman speculated that much of the high-frequency hearing loss in human ARHL arises from acoustic overstimulation [2].Intratympanic drug delivery to the inner ear is increasingly applied for both clinical and research purposes.However, the measurements of perilymph concentration achieved after intratympanic drug applications typically found the perilymph concentration to be substantially lower than the applied concentration (0.02-2.5 %).Even if there were no drug elimination from the eustachian tube and middle ear mucosa, it would take many hours for the apical drug levels to rise.Salt et al. compared the perilymph pharmacokinetics of steroids.As triamcinolone acetonide is lost rapidly from the perilymph, it will not spread far apically before being lost [29].Dexamethasone is distributed further along the cochlea but does not reach apical regions in appreciable concentrations.One-shot applications of aqueous solutions for a duration of a few hours result in steep drug gradients along the cochlea length, and limited drugs can reach the cochlea apex, which is especially important for low-frequency hearing loss [30].
The major factor controlling both the amount of drug entering the inner ear and the distribution of the drug along the cochlea length is the duration of the drug remaining in the middle ear space [20].In the present study, we combined poloxamers and MVLs to explore their sustained release performance and drug delivery capacity.According to the results, the mean diameter of ethosuximide-loaded MVLs was 14.7 μm, and more than 95 % of the particles were in the size range of 10-90 μm.The encapsulation efficiency of ethosuximide-loaded MVLs was determined to be 61.2 % ± 2.9 %.The gelling temperature was approximately 35.5℃, which was the closest to the normal ear temperature.Compared to the ethosuximide hydrogel encapsulating P407/ P188 (18 %/10 %) and ethosuximide-loaded MVL groups, the ethosuximide-loaded MVLs-Gel group showed the best sustained-release characteristics, up to over 120 h (5 days), with 73.3 % of ethosuximide released from the formulations.
It is well known that P407 can provide a depot for sustained release, and the numerous internal aqueous chambers of MVLs can provide high encapsulation efficiency and a longer duration of drug release.Animals in the intratympanic ethosuximide-loaded MVLs-Gel injection group showed significantly less SGN degeneration in the basal, middle, and apical turns compared to those in the intratympanic saline solution injection group.In particular, SGN loss was dramatically reduced in the apical turns (low-frequency regions) of mice treated with intratympanic ethosuximide-loaded MVLs-Gel injection compared to oral systemic ethosuximide administration.This indicates that ethosuximide can reach the apical turn of the cochlea through intratympanic sustainedrelease ethosuximide-loaded MVLs-Gel injection.
Consistent with the corresponding frequencies along the tonotopic axis of the mouse cochlea [31], intratympanic ethosuximide-loaded MVLs-Gel injection can significantly attenuate the ABR threshold shifts at all stimulus frequencies, including 4 kHz (low-frequency hearing), while oral systemic ethosuximide administration can significantly slow down age-related ABR threshold shifts at stimulus frequencies of 8, 16, and 32 kHz but not at 4 kHz.This indicates that ethosuximide can reach the apical turn of the cochlea through intratympanic ethosuximide-loaded MVLs-Gel injection but not through oral systemic ethosuximide administration.We hypothesized that P407 can provide a depot to continuously release encapsulated therapeutic molecules into the inner ear and that the numerous internal aqueous chambers of MVLs can provide high encapsulation efficiency and a longer duration of drug release.Consistent with this hypothesis, age-related SGN loss was dramatically reduced in the apical turns of mice treated with intratympanic ethosuximide-loaded MVLs-Gel injection compared to oral systemic ethosuximide administration.This indicated that ethosuximide-loaded MVLs-Gel increased the distribution of ethosuximide along the cochlea length.This is vital for SGNs in apical turns and corresponding low-frequency hearing.
In this study, with the combination of P407 and MVLs, we prepared a thermosensitive hydrogel containing ethosuximide-encapsulated MVLs.The ethosuximide-loaded MVLs-Gel showed high encapsulation efficiency and a longer duration of drug release.The ethosuximide-loaded MVLs-Gel increased the ethosuximide distribution along the cochlea length, especially the apical turn, which is vital for SGNs in apical turns and corresponding low-frequency hearing.Therefore, the use of ethosuximide-loaded MVLs-Gel provides a good solution for optimizing drug distribution in the cochlea.It is indicated that ethosuximide-loaded MVLs-Gel, as a novel drug delivery system, attenuates ARHL in C57BL/ 6J mice.Nevertheless, the underlying mechanism of Cav3.2 involved in ARHL remains unclear.It would be meaningful to explore the precise role of overexpression and knockout in the Cav3.2 mouse model.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig.1.Timeline of auditory brainstem response testing and drug delivery performed in this study (age in weeks).Intratympanic injections were repeated once every two weeks for a total of three times.Hearing was evaluated at the beginning (12 weeks of age) and at the end (20 weeks of age) of the study.

Fig. 2 .
Fig. 2. Ethosuximide-loaded MVLs at 400 × magnification under an optical microscope.Spherical with an internal appearance looked like aggregates of small particles.

Fig. 5 .
Fig. 5.After 8 weeks of treatment, typical images showed SGNs in Rosenthal's canal at basal turn (C,F,I), middle turn (B,E,H), and apical turn (A,D,G) of the intratympanic normal saline injection group (A,B,C), oral systemic ethosuximide-treated group (D,E,F), and intratympanic ethosuximide-loaded MVLs-Gel injection group (G,H,I).Age-related SGN loss was dramatically reduced in the apical turns of mice treated with intratympanic ethosuximide-loaded MVLs-Gel injection compared to oral systemic ethosuximide administration (p < 0.05).Data are given as mean ± SD (n = 10/group) and detected significances are marked with * (p < 0.05).

Table 1
Gelation temperature determination of different concentrations of P407 and P188 solutions.