The effect of noise exposure on insulin sensitivity in mice may be mediated by the JNK/IRS1 pathway

Five-week-old wild-type male ICR mice were obtained from the Qinglongshan Animal Center (Nanjing, China, SCXK(SU)2012-0008). In total, 144 mice were used in this study. The mice were housed in conventional cages with a 12-h light/dark cycle (lights on at 7 am) and had free access to food (standard rodent chow, SHOOBREE, Xietong Organism, Jiangsu, China) and water. After a 1-week acclimation period, the animals were randomly assigned into one control group or into one of three noise-exposure groups. The animals in the noise-exposure groups were exposed to broadband noise at a 95-dB SPL for 4 h/day between 8:00 am and 12:00 pm for 1 day (the N1D group), 10 days (the N10D group), or 20 days (the N20D group). The animals in each of the noise-exposure groups were further subdivided into three subgroups according to the times at which the assessments were performed: at 1 day, 1 week, or 1 month after the final noise exposure (1DPN, 1WPN, and 1MPN, respectively) (Fig. 1). The mice in the control group were also subdivided into the same subgroups according to the assessment time points and served as age-matched controls. The animals were treated humanely and with regard for the alleviation of suffering. Food consumption per cage was measured every 3 or 4 days by subtracting the amount of the food left from the initial amount of food supplied. To avoid circadian rhythm-induced variations, the insulin tolerance test (ITT) was always initiated at 9:00 am, and tissue and blood samples used for further studies were collected between 2:00 pm and 3:00 pm. All of the animal procedures were approved by the University Committee for Laboratory Animals of Southeast University, China (reference number: 20130307-004).

The animals were exposed to noise as described previously [4]. Briefly, after acclimatization to the noise-exposure setting for 30 min, awake and unrestrained mice were placed separately into metal net cages that were 50 cm below the horns of two loudspeakers. Electrical Gaussian noise generated by a System III processor from Tucker-Davis-Technologies (TDT, Alachua, FL, USA) was delivered to speakers after power amplification. In consistent with our previous studies, the acoustic spectrum of the sound was distributed mainly between 1 and 20 kHz [4, 19]. The noise level was monitored using a 1/4-in. microphone linked to a sound level meter (Larson Davis 824, Depew, NY, USA), and the sound intensity was maintained at a 95 ± 1 dB SPL.

The insulin tolerance test (ITT) on 4-h-fasted mice was initiated at 9:00 am to avoid circadian rhythm-induced variation. Blood samples were obtained from the tail vein of awake mice just prior to (0 min) and at 30, 60, and 90 min after an intraperitoneal injection of insulin (Humulin, 0.75 U/kg body wt). Blood glucose levels were measured using a portable glucose monitor (Bayer Contour, Bayer HealthCare LLC, Whippany, NJ) and test strips. The time course of absolute blood glucose recorded during the ITT and the areas under the blood glucose curves (AUC) were used to evaluate insulin sensitivity. After completion of the test, the mice were returned to their home cage and given free access to food and water.

Three to 4 h after completion of the ITT (i.e., 2:00–3:00 pm), the mice were decapitated 20 min after an intraperitoneal injection of insulin (Humulin, 0.75 U/kg body wt), and their gastrocnemius muscles were dissected and homogenized in ice-cold RIPA buffer (Beyotime P0013C, China) supplemented with a complete protease inhibitor cocktail (Roche, Germany) and PhosSTOP (Roche, Germany). The protein extracts (40 μg) for each preparation were separated using 10% SDS-PAGE and electrotransferred onto PVDF membranes (Millipore, Bedford, MA, USA). After blocking with Tris-buffered saline, 0.1% Tween 20, and 5% nonfat dry milk, the membranes were incubated with primary antibodies overnight at 4 °C. The following antibodies were used: anti-IRS1 (Cell Signaling Technology, Cat no. #2382, Beverly, MA, USA), anti-phospho-IRS1 (Ser307) (Cell Signaling Technology, Cat no. #2381, Beverly, MA, USA), anti-JNK (Cell Signaling Technology, Cat no. #9252, Beverly, MA, USA), anti-phospho-JNK (Thr183/Tyr185) (Cell Signaling Technology, Cat no. #9251, Beverly, MA, USA), anti-Akt (Cell Signaling Technology, Cat no. #4685, Beverly, MA, USA), anti-phospho-Akt (Ser473) (Cell Signaling Technology, Cat no. #4058, Beverly, MA, USA), and anti-phospho-Akt (Thr308) (Cell Signaling Technology, Cat no. #4056). The protein bands were visualized using an ECL Kit (WBKLS0050; Millipore, Billerica, MA, USA), and a densitometry analysis was performed using ImageJ.

Because insulin can influence the production of inflammatory cytokines and the oxidative stress response [20], separate groups of mice were used in the biochemical analyses. Blood was collected immediately from the trunk into dry tubes after the mice were decapitated without the pre-administration of insulin. Gastrocnemius muscles were then harvested, weighed, and homogenized to evaluate oxidative stress according to the protocols provided with assay kits for superoxide dismutase (SOD, a powerful endogenous enzymatic antioxidant) (STA-340, Cell Biolabs, Inc.), catalase (CAT, an important endogenous enzymatic antioxidant) (Cell Biolabs, STA-341, USA), and malondialdehyde (MDA, a major secondary oxidation product) (Cell Biolabs, STA-832, USA). Following centrifugation at 4 °C, plasma was separated from each sample, and plasma concentrations of TNF-a and IL-6 (Cat#EMC102a and Cat#EMC004, NeoBioscience, Shenzhen, China) were determined using enzyme-linked immunosorbent assay kits according to the manufacturer’s instructions and guidelines.

The food intake and the bodyweight of all groups tested at the three time points were similar (Fig. 2, inserts). At 1DPN, all noise groups exhibited blunted glucose responses to insulin challenge, as indicated by significantly higher blood glucose level(s) at one or more time point(s) during the ITT and larger AUC values (which were significant for N10D and N20D) compared with the control values (Fig. 2a, d). At 1WPN, the N1D group showed similar insulin sensitivity to that of the control group, and the N10D group only exhibited a significantly high blood glucose level at 30 min after insulin injections. However, the N20D group displayed not only significantly higher glucose levels at all three test points after insulin injections but also a significantly higher value of AUC (Fig. 2b, e). No differences were observed between any of the groups at 1MPN (Fig. 2c, f).

As demonstrated in Fig. 3, at 1DPN, the phosphorylation of JNK at Thr¹⁸³/Thr¹⁸⁵ and IRS1 at Ser³⁰⁷ were significantly elevated in all three noise-exposed groups, while the Akt phosphorylation at Thr³⁰⁸ and Ser⁴⁷³ induced by exogenous insulin stimulation exhibited decreases in all three noise-exposed groups, which reached significance in the N20D group for Thr³⁰⁸ phosphorylation and in the N10D and N20D groups for Ser⁴⁷³ phosphorylation. At 1WPN, significantly elevated phosphorylation of JNK was observed in the N10D and N20D groups, while the significantly elevated IRS1 phosphorylation and blunted Akt phosphorylation were only shown in the N20D group. No significant difference in phosphorylation levels from the controls was identified in any of the noise-exposed groups at 1MPN.

To explore whether inflammation and/or oxidative stress might be involved in the influence of noise exposure on insulin sensitivity, we examined the circulating levels of inflammatory cytokines (TNF-α and IL-6) and the tissue levels of oxidative stress markers (SOD and CAT activities and MDA concentrations) in mice subjected to 1 day and 20 days of noise exposure. Since no significant difference in the ITT and Western blotting assay was revealed between the groups at 1MPN, we did not perform these biochemical assays at 1MPN.

As illustrated by Fig. 4, the levels of the tested cytokines and oxidative stress markers were all comparable between the control and N1D groups at both 1DPN and 1WPN. However, the N20D group exhibited significantly higher plasma TNF-α and IL-6 levels at 1DPN and significantly elevated CAT activity and MDA concentrations in their skeletal muscles at both 1DPN and 1WPN, suggesting a transient elevation in systemic inflammatory responses and a prolonged oxidative imbalance in the skeletal muscle of the animals that were chronically subjected to repeated noise.