A total of 40 female guinea pigs (British white shorthair Dunkin, Hartley strain; age, 6–8 weeks; weight, 250–300 g) were obtained from the Military Academy of Medical Sciences Laboratory Animal Center (Beijing, China), where they were bred in-house. The quality of the guinea pigs had been tested by the State Food and Drug Administration Academy (Beijing, China) with laboratory animal production license no. SCXK, 2012–0004. The animals were housed in the Fengtai animal house of the Laboratory Animal Center of the Military Academy of Medical Sciences (Beijing, China). The environmental conditions including noise and cleanliness complied with national standards, with an indoor temperature of 18–22°C, 45–60% humidity and a 12 h light/dark cycle. Animals were fed guinea pig silage and fresh cabbage. The animal diet was verified prior to experimentation. Immunization was performed after 7 days. Each animal was weighed every week and observed for dietary intake, behavior and mental state. The animal experiment was approved by the People's Liberation Army (PLA) General Hospital and the Military Medical Science Animal Ethics Committee (Beijing, China). All procedures involving laboratory animals were performed according to the National Institutes of Health's Guiding Principles on the Care and Use of Animals.

Immunization of guinea pigs was performed according to the method of Nemoto et al (11). Purified rabbit myosin (5 mg/0.5 ml; cat. no. M1636), complete Freund's adjuvant (CFA; cat. no. F5881) and pertussis toxin (PT; cat. no. P7208; 50 µg/500 µl) were from Sigma-Aldrich (Merck KGaA; Darmstadt, Germany). Mycobacterium tuberculosis toxin (MTT; 600 mg inactivated bacteria, diluted with 5 ml normal saline, subjected to repeated freezing and thawing) was provided by the Tuberculosis Research Center of the 309th Hospital of the Chinese PLA (Beijing, China). To immunize the animals, 0.3 µl purified rabbit myosin (3 µg total; 5 mg/0.5 ml), 0.1 ml CFA and 10 µl MTT (1.2 mg total; 120 mg/ml) (14) were mixed with 0.2 ml normal saline. After thorough oscillation in a vortex mixer, the white emulsion was injected subcutaneously at multiple sites on the backs of guinea pigs. The injection was administered once a week for a total of 4 weeks. For the first 2 immune injections, each guinea pig was injected intraperitoneally with pertussis toxin at 500 ng/200 µl (14).

The EAM guinea pigs were randomly divided into the EAM/normal saline group (n=10), EAM/low-dose NBP group (n=10), and EAM/high-dose NBP group (n=10). NBP (1 g/ml) was provided by the Shiyao Pharmaceutical Co., Ltd. (Shijiazhuang, China). The NBP was dissolved in 0.5% Tween 80 (Beijing Keino Spring Biological Technology, Co., Ltd., Beijing, China) (15) and injected intraperitoneally into the EAM animals in the morning twice a week for 2 weeks at 40 µl/40 mg/kg wt (low dosage) or 80 µl/80 mg/kg wt (high dosage). For the normal saline group, the animals were injected with Tween 80 (5 g/kg wt) and normal saline (10 ml/kg wt).

The investigator was blinded to animal grouping. Rodent grasping power (as an indicator of strength) was determined using a grasping power apparatus (model no. KH-YLS13A; cat. no. 53007; maximum tension, 2,000 g; Beijing Kai Hui Shengda Technology Development Co., Ltd., Beijing, China) (16). The animal forelimbs were connected with a dynamometer, which records the maximum force on a wire grid (17). The range of power production is influenced by factors such as muscle inflammation, connective tissue changes, and spinal cord or brain neural plasticity (18). The tests were performed in triplicate and the tested maximum force (g) was regarded as the holding power.

Guinea pigs were anesthetized via intraperitoneal injection of 50 mg/kg sodium pentobarbital (Sigma-Aldrich; Merck KGaA) to prevent excessive struggle (19–26) A thoracotomy was performed and a needle was inserted into the right ventricle to harvest ~10 ml blood for use in the following experiments. Following blood harvest, guinea pigs skeletal muscle specimens were obtained from thigh muscles in the limbs of guinea pigs.

The sheared fresh muscle tissue was dissected (5 mm diameter, 1 cm length), and fixed using tragacanth gum. Approximately 1 l liquid nitrogen was placed into a heat insulation barrel, and a 100 ml beaker with a 30 cm iron wire handle was filled with 70–80 ml isopentane. The beaker was submerged in liquid nitrogen, and a specimen block was added and stirred continuously with large tweezers for 30 sec to freeze the specimen without ice crystals forming. The specimen was placed into liquid nitrogen for 30–60 min, until isopentane around the specimens had vaporized, and subsequently stored in sealed plastic bags at −80°C overnight. Specimens were subsequently heated to −25°C for 30 min and sliced into transverse sections (thickness, 5 µm) using a sharp blade. Slices were stored at −25°C for staining. A portion of the muscle tissue from the distal and proximal limbs of each of the 40 guinea pigs was prepared as a frozen section, with 8 sections per guinea pig. Fresh muscle tissues (50–100 mg) were immediately placed in sterilized cryopreserved tubes, immersed in liquid nitrogen and stored at −80°C for subsequent analysis using polymerase chain reaction.

Samples were stained in Harris hematoxylin solution for 10 min, rinsed in water for 10 min, stained in 1% eosin for 3 min, and then repeatedly rinsed in water. After dehydration in ethanol, the samples were rinsed in xylene and coated with balata. Pathological scores were determined by observing 4 visual fields per section under a light microscope (Olympus-BX43; Olympus Corporation, Tokyo, Japan). Scoring of inflammation was performed using the following classification method (27,28): 1, ≤5 muscle fibers involved; 2, damage involving 5–30 muscle fibers; 3, damage involving 1 muscle bundle; 4, diffuse, extensive damage, involving >1 muscle bundle. For muscles with multiple-site damage, an additional 0.5 points were added to the score.

Muscle tissue blocks (0.2–1.0 g) were rinsed in cold normal saline, dried with filter paper and weighed. The blocks were transferred to 10-ml Eppendorf (EP) tubes, cut into small pieces and immediately transferred to a small mortar with ice water. Nine volumes of pre-cooled 0.9% sodium chloride were used to flush residual broken tissue from the EP tubes to obtain a suspension containing 10% muscle tissue homogenate. The samples were ground repeatedly and adequately with a pestle (10–15 min) to prepare a tissue homogenate. The muscle tissue homogenates were transferred to EP tubes, centrifuged at 4°C and 350 × g for 5 min, and the supernatant was collected.

After centrifugation of the above mentioned 10% tissue homogenate (4°C, 700 × g, 10 min) and sedimentation of cell debris and minor tissue, the mitochondria were collected via centrifugation at 4°C (3,500 × g for 15 min). The protein concentrations among different samples were standardized by Coomassie staining and spectrophotometry [optical density at 595 nm (OD595) was measured].

Specific steps were taken according to the kit instructions (ATPase kit; Nanjing Jiancheng Bioengineering Institute, Nanjing, China). One unit of ATPase enzyme activity [µmol inorganic phosphorus (pi)/mg protein/hour] is defined as the amount of ATPase that catalyzes the production of 1 µmol pi per hour per mg protein. ATPase activity was calculated from the OD using the following formula: ATPase activity = [(OD_sample – OD_control) / OD_standard] × 16.8.

RNA was extracted with TRIzol and the concentration was determined spectrophotometrically. Reverse transcription was performed using the RevertAid First Strand cDNA Synthesis kit (Thermo Fisher Scientific, Inc.). Total RNA (1 µg), random primers (1 µl), 5X Reaction Buffer (4 µl), Ribolock RNase inhibitor (20 U/µl; 1 µl), 10 mM dNTP mix (2 µl), RevertAid M-MuLv reverse transcriptase (200 U/µl; 1 µl) and nuclease-free water (added to reach a final reaction volume of 20 µl) were combined into RNAse-free tubes on ice. The reaction was gently mixed, centrifuged (350 × g for 1 min at room temperature to avoid blister occurrence) and incubated at 25°C for 5 min, 42°C for 60 min and 70°C for 5 min. The reverse transcription product was used directly for polymerase chain reaction (PCR), stored at −20°C for <1 week, or kept at −70°C for long-term storage.

The following primers were used: GAPDH forward, 5′-GCCGCATCGGTATTCCTTCT-3′ and reverse, 5′-GCGTCCAATACGGCCAAATC-3′; interferon (IFN)-γ forward, 5′-CACCATCTGGTTGCTGCCTA-3′ and reverse, 5′-TACCAGGGTACCAGGCCATT-3′; retinoic acid receptor-related orphan nuclear receptor (ROR)γt forward, 5′-TGATAGGTGGATTTGCGGGA-3′ and reverse, 5′-GCTGGGCCCAAAGCTAAAGT-3′; and forkhead box (Fox)p3 forward, 5′-TCCACAATGGGACTCATGCC-3′ and reverse, 5′-CTGTGGAGAGCTGGTGCATA-3′ (all Sangon Biotech, Shanghai, China).

Contents of the PCR mixture were as follows; 12.5 µl 2X qPCR Master Mix, 0.5 µl each of forward and reverse primers (0.3 µM), 1 µl cDNA (500 ng) and nuclease-free water to make the volume up to 25 µl. The contents were mixed and packaged into PCR tubes before being centrifuged for 1 min at 350 × g at room temperature to avoid blister occurrence.

PCR was performed in an iQ5 real-time fluorescent quantitative PCR cycler (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The PCR thermocycling conditions were as follows: Initial denaturation at 95°C for 10 min, annealing at 95°C for 15 sec and denaturation at 60°C for 30 sec for 40 cycles. The extension step was performed at 72°C for 30 sec followed by 80 cycles of final extension at 55–95°C for 30 min to create a dissociation curve. PCR products were quantified using the 2^−∆∆Cq method (29) and GAPDH was used as a reference gene.

Values are expressed as the mean ± standard deviation. Depending on the parameter and comparison, one-way analysis of variance and the least-significant differences test as well as Dunnett's t-tests were performed using SPSS 21.0 statistical software (International Business Machines, Armonk, NY, USA).

To establish an animal model of idiopathic inflammatory myopathy, 6- to 8-week-old guinea pigs were obtained and monitored for 7 days. Two out of 40 guinea pigs failed to adapt to their environment, as they ate less and lost weight. However, after another week, nutrition and weight were restored, which suggested successful adaptation to the environment for formal testing.

The guinea pigs were randomly assigned to a normal control group and three EAM model groups (n=10 per group). Guinea pigs in the normal control group showed normal eating and general behavior and had a healthy skin color (Fig. 1A). Guinea pigs in the EAM model group developed redness, swelling and induration (0.4–1.0 cm in diameter) at the injection site after the second week (Fig. 1B). The animals became listless, fearful and thin. The skin started to depilate, the weight gain decreased slightly, movement slowed and the cry of the guinea pigs became low or even hoarse. The symptoms grew progressively more debilitating with time. The body weight ceased to increase or began to decrease in the fourth week, the hair coat became rough and lacklustre, the abdomen and buttocks appeared dirty, signs of malnutrition were apparent, and the guinea pigs developed myasthenia in both forelimbs and were unable to hold up their extremities (Fig. 1C). One guinea pig had severe myasthenia and lay laterally (Fig. 1D). The animal was immediately euthanised by exsanguination under anesthesia (intraperitoneal injection of 15 mg/kg sodium pentobarbital; Sigma-Aldrich; Merck KGaA). Simultaneously, blood samples and muscle specimens were collected.

To determine the effects of NBP on guinea pigs with EAM, the animals with induced EAM were treated with normal saline and low- or high-dose NBP (n=10 guinea pigs per group). In guinea pigs with EAM treated with low-dose NBP, the weight was not obviously affected. The EAM/high-dose NBP group had more weight gain than the normal saline group, but the differences were not significant (Fig. 2A). However, the guinea pigs in the EAM/normal saline group had significantly lower strength than those in the control group (P=0.036; Fig. 2B). The guinea pigs with EAM treated with MBP had a higher strength than those in the EAM/normal saline group, with the difference being significant in the EAM/high dose NBP group (P=0.045). These results suggested that NBP partially reversed the myasthenia caused by EAM.

To determine the effects of NBP at the microscopic level, muscle tissues were subjected to histopathological analysis. The degree of inflammatory cell infiltration was low in control guinea pigs (Fig. 3A), but was clearly increased upon EAM induction (Fig. 3B). EAM guinea pigs treated with low-dose NBP (Fig. 3C) showed a slight reduction in the inflammatory cell infiltration, while EAM guinea pigs treated with high-dose NBP (Fig. 3D) showed a marked reduction in inflammatory cell infiltration. The levels of inflammation in the 4 groups of guinea pigs were scored by selecting 4 visual fields in each of the 8 sections per guinea pig (n=10 guinea pigs per group). Compared with the normal control group, the EAM group had a significantly increased pathological inflammation score (P<0.01). Compared with the EAM/normal saline group, the EAM/high dosage NBP group had a significantly reduced pathological inflammation score (P<0.05). The EAM/low dose-NBP group also showed a reduction, which was not significant (Fig. 4).

To verify the abovementioned findings at the molecular level, the Ca²⁺-ATPase activity of the muscle mitochondrial membrane and muscle plasma membrane was measured. Each of these measures of muscle activity was significantly decreased upon EAM induction. However, this decrease was significantly attenuated in the EAM/high-dose NBP group compared with the EAM/normal saline group (P<0.05; Fig. 5), confirming that NBP reverses the effects of EAM on muscle activity. In the EAM/low-dose NBP group, Ca²⁺-ATPase activity in the muscle mitochondrial membrane was also significantly increased compared with the EAM/saline group (P<0.05; Fig. 5).

To determine the effects of NBP on T-cell-associated cytokines, the mRNA expression levels of IFN-γ, RORγt and Foxp3 in muscle tissues were measured. The expression of IFN-γ mRNA was significantly reduced by NBP (P<0.01; Fig. 6A), whereas Foxp3 and RORγt mRNA expression levels were significantly elevated by NBP (P<0.01; Fig. 6B and C).