A total of 60 female C57BL/6 mice (age, 6 weeks) with a body weight of 20±2 g were purchased from Vital River Laboratories Co., Ltd., (Beijing, China), provided with free access to food and water and raised in sterilized conditions for 1 week at 24±2°C in a humidity of 40–50% and a 12 h dark/light cycle. A total of 10 C57BL/6 mice (male, n=5; female, n=5; age, 1 week; weight, 4–5 g) used for the cultivation of primary MSCs were purchased from the Animal Center of the Academy of Military Medical Sciences (Beijing, China), and housed in the same conditions described above. C3H10T1/2 cells were purchased from the American Type Culture Collection (Manassas, VA, USA). The C3H10T1/2-MIGR1/MSC and C3H10T1/2-MIGR1-ICAM-1/MSC cell lines were constructed in the authors' laboratory at Tianjin Medical University General Hospital (Tianjin, China). Study protocols were approved by the Ethics Committee of Tianjin Medical University General Hospital (Tianjin, China).

Porcine thyroglobulin, complete Freund's adjuvant and incomplete Freund's adjuvant were purchased from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). Commercial ELISA kits for total triiodothyronine (TT3; cat. no. 80985; Crystal Chem, Inc., Downers Grove, IL, USA), total thyroxine (TT4; cat. no. 80983; Crystal Chem, Inc.), thyroid stimulating hormone (TSH; cat. no. RTC700R; Xinbosheng Biotech Co., Ltd., Shenzhen, China) and anti-thyroid peroxidase (TPOAb; cat. no. 05–50085; Zhuzhou Zeye Biotech Co., Ltd., Zhuzhou, China), anti-thyroid microsomal (TMAb; cat. no. KB12639; Shanghai Jianglai Biotech Co., Ltd., Shanghai, China) and anti-thyroglobulin antibodies (TGAb; cat. no. MA512408; Beinuo Life Science, Shanghai, China) were used. TRIzol reagent was purchased from Gibco (Thermo Fisher Scientific, Inc., Waltham, MA, USA). TaqDNA polymerase, dNTP mixture, PrimeScript reverse transcription kit (cat. no. 639505), RNase inhibitor and oligo d(T)15 primers were purchased from Takara Bio, Inc., (Otsu, Japan). Rabbit anti-mouse antibodies against total-mitogen-activated protein kinase p38 (p38; cat. no. 9212), phosphorylated (p)-P38 (cat. no. 9216), total extracellular signal-regulated kinase (ERK) (cat. no. 9101) and p-ERK (cat. no. 4696), and goat anti-rabbit horseradish peroxidase (HRP)-conjugated antibodies were purchased from Cell Signaling Technology, Inc., (Danvers, MA, USA). Anti-mouse β-actin antibodies (cat. no. 3700) were also purchased from Cell Signaling Technology, Inc., and CM-DiI was purchased from Sigma-Aldrich (Merck KGaA). Type II collagenase, α-MEM medium and fetal bovine serum were purchased from Gibco (Thermo Fisher Scientific, Inc.).

The EAT model was established in 6-week-old C57BL/6 mice according to a previously described method (13). Subsequently, mice were randomly divided into the following groups (n=10 each): Normal control; EAT model; primary MSC group, which was subjected to EAT induction, followed by administration of MSCs (3×10⁵) via the caudal vein; C3H10T1/2/MSC group with similar function of MSCs, which was subjected to EAT induction, followed by administration of C3H10T1/2/MSCs (3×10⁵) via the caudal vein; C3H10T1/2-MIGR1/MSC group, which was subjected to EAT induction, followed by administration of C3H10T1/2-MIGR1/MSCs (3×10⁵) via the caudal vein; and C3H10T1/2-MIGR1-ICAM-1/MSC group, which was subjected to EAT induction, followed by administration of C3H10T1/2-MIGR1-ICAM-1/MSCs (3×10⁵) via the caudal vein. C3H10T1/2-MIGR1/MSCs were cells transfected with C3H10T1/2-MIGR1 plasmid, and C3H10T1/2-MIGR1-ICAM-1/MSC were cells transected with C3H10T1/2-MIGR1-ICAM-1 plasmid.

Blood samples were harvested from the angular vein of each mouse at day 28 post-treatment to determine the serum TT3, TT4, TSH, TPOAb, TMAb and TGAb according to the manufacturer's protocol. All tests were performed a minimum of three times.

Following anesthesia with 3% pentobarbital (Sigma-Aldrich; Merck KGaA), mice were sacrificed via cervical dislocation. An incision was subsequently made on the skin along the cervical median, and the muscles were separated to expose the thyroid gland. The gland was removed using an ophthalmic scissor. Thyroid gland tissues were fixed using 4% formalin and embedded in wax, followed by preparation of 4 µm-thick sections. H&E staining was subsequently performed to evaluate thyroid gland injury by evaluating the size of the follicles, inflammatory infiltration, colloidal substance retention and epithelial injury.

Following sacrifice, an incision was made under the inferior margin of the left rib to expose the spleen. The spleen was obtained using an ophthalmic scissor. The width, thickness and craniocaudal length of the spleen were determined and were used to calculate the splenic index according to the previous description (14).

Total RNA was extracted from splenic cells using TRIzol reagent according to the manufacturer's protocol. cDNA synthesis was performed using ~2 µg RNA with PrimeScript reverse transcription kit qPCR was performed using SYBR-Green on an Applied Biosystems 7500 Software v2.0.6 system (Thermo Fisher Scientific, Inc., Waltham, MA, USA) with the primers listed in Table I. PCR reactions were performed in a total volume of 10 µl containing 5 µl 2X SYBR Premix, 0.2 µl each specific primer to a final concentration of 200 nM, and 1 µl cDNA template. The PCR conditions consisted of denaturation at 94°C for 3 min, followed by 40 cycles of 94°C for 30 sec, 60°C for 30 sec, and 72°C for 60 sec. The mRNA level was normalized to that of β-actin, and all reactions were performed at least in triplicate. The amplification results for real-time PCR were calculated using the 2^−ΔΔCq method (15).

Splenic tissues were homogenized at room temperature for 2 min in radioimmunoprecipitation assay buffer containing protease and phosphatase inhibitors (BB-3201-1; Bestbio, Shanghai, China). Protein concentration was determined using a bicinchoninic acid assay, and 40 µg protein per lane was separated by 10% SDS-PAGE and transferred to a Hybond-P membrane. The membrane was subsequently blocked with 5% non-fat milk and incubated with rabbit anti-mouse p-p38 (1:200), total-p38 (1:500), p-ERK (1:1,000) and total-ERK (1:200) antibodies overnight at 4°C. Subsequently, the mixture was incubated with the HRP-conjugated goat anti-rabbit antibody (1:1,000) for 1 h at room temperature. Following washing with Tris-buffered saline Tween-20, the bound primary antibody was visualized using enhanced chemiluminescence and exposed to X-ray film. The relative density of each protein was analyzed using TL120 imaging software (version v2009; TotalLab, Ltd., Newcastle upon Tyne, UK). The same membrane probed with anti-β-actin antibodies (1:2,000) served as the loading control. All experiments were performed a minimum of three times.

To determine the migration of MSCs in vivo, 1-week-old C57BL/6 mice were randomly divided into the following groups, subsequent to EAT induction (n=10 each): i) CM-DiI-labeled primary MSCs; ii) CM-DiI-labeled C3H10T1/2/MSC; iii) CM-DiI-labeled C3H10T1/2-MIGR1/MSC; and iv) CM-DiI-labeled C3H10T1/2-MIGR1-ICAM-1/MSC, which were administered the relevant cells via the caudal vein. For the CM-DiI labeling process, MSCs (2×10⁶) were resuspended with 1 ml PBS and labeled with 2 µl CM-DiI at 37°C for 5 min. Subsequently, the mixture was incubated at 4°C for 5 min, and centrifuged at 179 × g for 10 min.

Following anesthesia with 3% pentobarbital (Sigma-Aldrich; Merck KGaA) and sacrifice via cervical dislocation, mice were fixed on an operating table to expose the chest, the chest was shaved and the fascia removed, and the sternum and ribs were removed to expose the heart. PBS was infused from the left ventricle to expel circulated blood from the right atrium. Subsequently, the lung and thyroid gland were washed using PBS, and fixed with 10% paraformaldehyde. Following embedding, 5-µm sections were cut and observed via fluorescence microscopy to determine the positive staining.

SPSS 18.0 software (SPSS, Inc., Chicago, IL, USA) was used to perform statistical analysis. Data are presented as the mean ± standard deviation, and Student's t-test was performed for inter-group comparisons. P<0.05 was considered to indicate a statistically significant difference.

No statistical difference was observed in the TT3 levels between groups (Fig. 1A). For the expression of TT4, a significant decrease was observed in the EAT model group compared with that of normal controls (P<0.05). However, following administration of MSCs, the level of TT4 was markedly increased compared to the model group. Compared with the C3H10T1/2-MIGR1/MSC group, the level of TT4 was significantly increased in the C3H10T1/2-MIGR1-ICAM-1/MSC group (P<0.05; Fig. 1B). The expression of TSH significantly decreased in the C3H10T1/2-MIGR1-ICAM-1/MSC group compared with the C3H10T1/2-MIGR1/MSC group (P<0.05; Fig. 1C). Levels of TPOAb, TGAb, TMAb and TSH were significantly higher in the EAT model group compared with those of normal controls (all P<0.05; Fig. 1C-F). Similarly, following administration of MSCs, their expression was decreased, particularly in the C3H10T1/2-MIGR1-ICAM-1/MSC and MSC groups. Compared with the C3H10T1/2-MIGR1/MSC group, the levels of TPOAb, TGAb, TMAb and TSH were significantly decreased in the C3H10T1/2-MIGR1-ICAM-1/MSC group (all P<0.05).

Compared with the control group (Fig. 2A), injury of thyroid follicles was determined visually by their uneven size and the presence of epithelial hyperplasia and infiltration of inflammatory cells in the EAT model group (Fig. 2B) as revealed by H&E staining. In the C3H10T1/2-MIGR1-ICAM-1/MSC and MSC groups thyroid gland injury, epithelial hyperplasia and infiltration of inflammatory cells were attenuated;. however, in the other groups, no marked difference was observed following the administration of MSCs (Fig. 2C-F).

The splenic index in the EAT model group was significantly higher than that of the normal control group (P<0.01). In the C3H10T1/2-MIGR1-ICAM-1/MSC group, the splenic index was signifantly decreased compared with the EAT model group (P<0.05). However, no significant improvement was observed in the splenic index in the other treatment groups (Fig. 3).

RT-qPCR indicated that IL-4 and IL-10 mRNA expression was downregulated in the EAT model group compared with that of normal controls (P<0.05; Fig. 4A and B). However, this downregulation was reversed in the C3H10T1/2/MSC (IL-4, P<0.01), C3H10T1/2-MIGR1/MSC (IL-4, P<0.05; IL-10, P<0.01), C3H10T1/2-MIGR1-ICAM-1/MSC (IL-4 and IL-10, P<0.01) and MSC groups (IL-4, P<0.05; IL-10, P<0.01) compared with the model group. The mRNA expression of IL-17 and INF-γ was significantly increased in the EAT model group compared with normal controls (both P<0.01; Fig. 4C and D). This upregulation was significantly attenuated following treatment with MSCs (P<0.05). The upregulation of IL-4 and IL-10 mRNA and downregulation of IL-17 and INF-γ mRNA were most apparent in the C3H10T1/2-MIGR1-ICAM-1/MSC group compared with the other MSC groups.

Western blot analysis revealed the expression of p-ERK and p-p38 was upregulated in the model group compared with the normal control group. Compared with the model group, the expression of p-ERK and p-p38 was decreased in the C3H10T1/2-MIGR1-ICAM-1/MSC group. No statistical difference was observed in the expression of p-ERK and p-p38 in the C3H10T1/2-MIGR1/MSC, C3H10T1/2/MSC and primary MSC groups compared with the model group (Figs. 5 and 6).

Red fluorescence was observed in the thyroid gland (Fig. 7) and lung (Fig. 8), indicating the presence of CM-DiI-labeled MSCs. Localization in these tissues was markedly increased in the C3H10T1/2-MIGR1-ICAM-1/MSCs group in the thyroid gland compared with the other groups. The localization of primary MSCs in the lung was superior to that in the thyroid gland in all groups, which may be associated with the arrest of MSCs by the extensive vascular network in the lung.