Because tinnitus is most prevalent in older humans, we used the SAMP8 mice for this study, although other strains of rodents, cats, or rabbits could also have been used. Forty-eight 3-month-old male SAMP8 mice, weighing between 22 and 33 g, were randomly and equally divided into two groups: salicylate-treated (n = 24) and saline-treated (n = 24). Our Institutional Animal Care and Use Committee approved the protocols used in this study.
Behavioral measurement of tinnitus score
All mice were trained to perform an active avoidance task, which was performed in a conditioning box with an electrical floor and a climbing pole, according to the design of Guitton et al. .
Conditioning to the task
The conditioning paradigm consisted of 6 sessions performed daily for 5 days (day 1 to 5). Each session lasted 15 to 20 min, and there were 10 trials per session. Inter-trial intervals were at least 1 minute. For each trial, the conditioning stimulus was a 50 dB sound pressure level (SPL) pure tone with a frequency of 10 k Hz and a 3-second duration. The unconditioned stimulus was a 3.7 mA electric foot-shock presented for up to 30 seconds, as described by Guitton's protocol , by adjusting the electric voltage with fixed copper wire resistance on the floor. The time between the conditioned stimulus and the unconditioned stimulus was 1 second. The mice would climb up the pole to a safe area after the coupled conditioned and unconditioned stimuli. Electrical shocks were stopped by the experimenter when the animal climbed correctly. The "true-positive" score was the level of performance as assessed by the number of times the mice climbed correctly in response to sound. Mice were considered to be conditioned when the "true-positive" score reached at least 80% in three consecutive sessions. Only conditioned mice entered the tinnitus experiments.
Induction and testing of tinnitus
Once conditioned, the mice rested for 1 day (day 6). Then, for 4 consecutive days (days 7 to 10), an active avoidance task was performed 2 hours after intraperitoneal injections of saline either alone or containing 300 mg/kg sodium salicylate (Sigma, St. Louis, MO) [1, 2, 20]. The active avoidance task consisted of one session with 10 trials. To compensate for hearing loss (threshold elevation of 15 to 20 dB SPL during 4-days injections in our preliminary, unpublished data) induced by salicylate injection, the intensity of sound that elicited the behavioral responses was increased to 70 dB SPL for the salicylate-treated group. By doing so, the sound sensation levels of all mice in both groups were similar.
During testing, a sound of 3-second duration was given and the mice were observed for another 5 seconds to see whether they would climb to the safe area as conditioned (true-positive). If the animals stayed in the safe area ≥10 seconds, the mice were returned to the floor for ongoing observation. If the mice did not climb to the safe area in response to the sound, an electrical shock was given by the experimenter to remind the mice to climb to the safe area. Again, if animals stayed in the safe area ≥10 sec, the mice were returned to the floor for ongoing observation. Finally, the experimenter observed the total number of times (tinnitus score) the mice climbed during the inter-trial silent period of 1 minute (false-positive climbs) for 10 trials.
Samples isolation and RNA extraction from cochlea and IC
The body weights of all mice were measured at day 7 before salicylate or saline injection, and at day 10 before the mice were euthanized by decapitation. The weights of the midbrain including IC were also measured before dissection. Paired samples of cochlea and IC were immediately dissected using a Zeiss stereomicroscope and stored in a -80°C freezer until use. RNA isolation was performed using RNA-Bee isolation reagent (Friendswood, USA) with a tissue homogenizer according to the manufacturer's protocol. RNA quality was assessed using an Agilent Bioanalyzer 2100 and the ratio of absorbance measurements at 260 and 280 nm.
Reverse transcription-polymerase chain reaction (RT-PCR)
High quality RNA was used as substrate to synthesize cDNA by reverse transcription using a MasterAmp™ High Fidelity RT-PCR Kit (Epicentre Biotechnologies, USA) in a P × 2 Thermal cycler (Thermo Electron Corporation Bioscience Technologies Division, USA). RT was carried out at 37°C for 1 hour. For PCR amplification, 7.5 μL cDNA and primers were used according to the supplier's instructions. The primers were: TNF-alpha-F, 5'-CCCCTCAGCAAACCACCAAG-3', TNF-alpha-R, 5'-CTTGGCAGATTGACCTCA GC-3'; IL-1beta-F, 5'-GAGTGTGGATCC CAAGCAAT-3', IL-1beta-R, 5'-CTCAGTGCAGGCTATGACCA-3'; NMDA receptor subtype 2B (NR2B)-F, 5'-TCC GCC GAG AGT CCT CCG T-3', NR2B-R, 5'-CTG CGT TGC CCT CGA TGT T-3'; β-actin-F, 5'-CCACACCCGCCACCAGTTCG-3', and β-actin-R, 5'-CCCATTCCCACCATCACACC-3' (Protech-taiwan, Taiwan).
The thermal cycling conditions for PCR were adjusted to the following: 3 min initial set-up at 95°C; followed by 50 cycles of denaturation (45 s at 95°C for all genes), annealing (45 s at 53°C for TNF-alpha, 52°C for IL-1beta, 54°C for NR2B, and 50°C for β-actin. A final 10 min extension at 72°C was included for all thermal cycle runs.
Quantification of PCR products
The DNA products were measured using a Mini Horizontal Electrophoresis System (MJ-105/MP-100, Major Science, Taiwan) and an E-Box-1000/26M Inspection Certificate & Analysis Systems (E-Box Spp-010 E-capt soft ware, USA). The expression levels of TNF-α, IL-1β and NR2B genes are presented as relative ratios in comparison to β-actin.
Data are presented as mean ± standard deviation (SD), unless indicated otherwise. Expression levels for the TNF-α or IL-1β genes, presented as ratios relative to β-actin, were compared for both groups using Student's t-test. The correlation between the tinnitus score (dependent variable) and TNF-α or IL-1β gene (independent variable) expression, and between NR2B (dependent variable) and TNF-α or IL-1β gene (independent variable) expression, were analyzed using a linear regression model. All of the above analyses were performed using the commercialized software "STATA10", and p values < 0.05 were considered statistically significant.