Synthetic decoy ODNs were synthesized by Macrogen, Inc. using the following ODN sequences (the target sites of the consensus-binding sequences are underlined): STAT3 decoy ODN, 5′-GAA TTC CTT CTG GGA ATT CCA AAA GGA ATT CCC AGA AG-3′; NF-κB decoy ODN, 5′-GAA TTC AGG GAAATC CCT TCA AGA AAA CTT GAA GGG ATT TCC CT-3′; and scramble (Scr) ODN, 5′-GAA TTC AAT TCA GGG TAC GGC AAA AAA TTG CCG TAC CCT GAA TT-3′. Subsequently, STAT3, NF-κB, STAT3/NF-κB decoy ODNs and Scr decoy ODN were annealed for 6 h, decreasing the temperature from 80 to 25°C during this time. To obtain a covalent ligation for ring-type decoy ODNs, each ODN was mixed with T4 DNA ligase (Takara Bio, Inc.) and incubated for 18 h at 16°C. These decoy ODNs were predicted to form a covalently ligated ring-type structure (Fig. 1A).

A total of 60, male C57BL/6 mice (age, 6 weeks; weight, 20-25 g; Samtaco Bio Korea Co., Ltd.) were housed in a room at a controlled temperature of 22±2°C and a humidity of 55%, under a 12-h light/dark cycle. The mice were given free access to water and food. Atherosclerotic injuries were induced via intraperitoneal injection of LPS from Escherichia coli O111:B4 (2 mg/kg body weight; dissolved in 200 µl PBS; MilliporeSigma) once a week for 8 weeks. Simultaneously, the mice were fed an atherogenic diet (AD; 21.2% milkfat, 1.25% cholesterol, 0.5% cholic acid; DooYeol Biotech.). Decoy ODNs (10 µg) were administered every 2 weeks for 8 weeks via an injection into the tail vein using a Trans IT in vivo gene delivery system (Mirus Bio.). The mice were divided into the following six groups (n=10 mice/group): i) Untreated group [normal control (NC)], ii) STAT3/NF-κB decoy ODN-treated group (STAT/NF-κB) fed a standard diet (cat. no. 2018S; Envigo), iii) Scr decoy ODN-treated group injected with LPS and fed an AD (LPS + AD + Scr), iv) STAT3 decoy ODN-treated group injected with LPS and fed an AD (LPS + AD + STAT), v) NF-κB decoy ODN-treated group injected with LPS and fed an AD (LPS + AD + NF-κB), vi) STAT3/NF-κB decoy ODN-treated group injected with LPS and fed an AD (LPS + AD + STAT/NF-κB). All mice were anesthetized with 2-3% isoflurane inhalation (Ifran; HANA Pharm Co., Ltd.) using an RC2 Rodent Circuit Controller (VetEquip, Inc.) before tail vein injection. At the end of each treatment period, blood was collected by cardiac puncture from the mice, and the mice were euthanized by asphyxiation with 60-70% CO₂ of the cage volume/min. Subsequently, their aorta and heart tissues were excised for subsequent experiments. The humane endpoints in the present study were as follows: i) The animals showed hypothermia (<37°C) in the absence of anesthesia; ii) the animal lost 20% of its original weight; iii) the animal exhibited loss of ability to ambulate (unable to access food or water). All animals reaching these endpoints were euthanized with 60-70% CO₂ of the cage volume/min. Death verification included the absence of heartbeat, breathing or respiration. All animal protocols were approved by the Institutional Animal Care and Use Committee of the Catholic University of Daegu (Daegu, South Korea; approval no. DC IAFCR-181204-27Y). All procedures performed in experiments involving animals were in accordance with the ethical standards of the institution or practice at which the studies were conducted.

Nuclear extract fractionation was performed on the abdominal aorta tissues of mice using an NE-PER™ Nuclear and Cytoplasmic Extraction Kit (Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions based on standard protocols. Subsequently, a Lightshift Chemiluminescent EMSA kit (cat. no. 20148; Thermo Fisher Scientific, Inc.) was performed to analyze STAT3 and NF-κB DNA binding activation. ODNs containing the consensus STAT3 and NF-κB binding sites (STAT3, forward 5′-GAT CCT TCT GGG AAT TCC TAG ATC-3′, reverse 5′-GAT CTA GGA ATT CCC AGA AGG ATC-3′; NF-κB, forward 5′-CTT GAA GGG ATT TCC CTG GCT-3′, reverse 5′-AGC CAG GGA AAT CCC TTC AAG-3′) and 3′ end-labeled with biotin were used as probes. Biotin-labeled probes were synthesized by Macrogen, Inc. The probes (20 fmol) were subjected to a denaturation step at 90°C for 1 min followed by annealing at room temperature for 30 min. Then, for the binding reaction, the binding reaction components were added in the order listed according to the manufacturer's instructions. At this time, 10 µg nuclear extract was added to binding reactions and incubated for 20 min at room temperature. The nuclear protein concentration (1-10 µg/µl) was determined using the Bradford protein assay (Bio-Rad Laboratories, Inc.). The Novex™ TBE gels 4-12% (cat. no. EC62352; Thermo Fisher Scientific, Inc.) were used for pre-electrophoresis for 30-60 min at 100 V. The sample complexes were separated by electrophoresis on gels using 0.5X TBE (cat. no. LC6675; Thermo Fisher Scientific, Inc.) as a running buffer until the bromophenol blue dye has migrated ~3/4 down the length of the gel. After electrophoresis, the gels were transferred to nylon membranes and detected using the Streptavidin-Horseradish Peroxidase Conjugate and the Chemiluminescent Substrate (cat. no. 89880; Thermo Fisher Scientific, Inc.), according to manufacturer's instructions. The signal intensity was detected using an image analyzer (ChemiDoc™XRS+; Bio-Rad Laboratories, Inc.).

All abdominal aorta and heart tissue specimens were fixed in 10% formalin for 24 h at room temperature. Thereafter, sections of the aorta and heart were dehydrated in graded ethanol, cleared in xylene and embedded in paraffin. The sections (4 µm) were mounted on glass slides, rehydrated in distilled water and stained with hematoxylin and eosin (H&E), Verhoeff-Van Gieson and Masson's trichrome based on standard protocols. The tissue sections were deparaffinized and stained with hematoxylin at room temperature for 8 min and then with eosin at room temperature for 5 min. For the Verhoeff-Van Gieson stain, the tissue sections were deparaffinized and stained with alcoholic hematoxylin, 10% ferric chloride and Weigert's iodine mixed solution for 5 min at room temperature. After differentiating in 2% ferric chloride for 1-2 min and treating with sodium thiosulfate for 1 min, the sections were counterstained with Van Gieson's solution for 40 sec. For the Masson's trichrome stain, the tissue sections were deparaffinized and refixed in Bouin's solution for 30 min at 60°C. After being stained at room temperature with Weigert's hematoxylin and a Biebrich scarlet-acid fuchsin each for 10 min, the sections were treated in phosphomolybdic–phosphotungstic acid for 10 min and finally stained with aniline blue for 10 min. All slides were examined under a Pannoramic MIDI slide scanner (3DHISTECH Kft.).

The aorta tissue specimens were lysed in protein lysis buffer (CelLytic™ M; MilliporeSigma) for 20 min on ice and were then centrifuged at 13,800 × g for 20 min at 4°C. Nuclear and cytosolic protein samples were prepared from the aorta tissue using NE-PER Nuclear and Cytoplasmic Extraction Kit (Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. The supernatant was collected and the protein concentration was measured using the Bradford protein assay (Bio-Rad Laboratories, Inc.). The protein samples were then loaded on precast gradient SDS polyacrylamide gels (Bolt™ 4-12% Bis-Tris Plus Gels; Thermo Fisher Scientific, Inc.) and transferred to nitrocellulose membranes (GE Healthcare) using a Bolt™ Mini Blot Module and Mini Gel Tank (Thermo Fisher Scientific, Inc.), according to the manufacturer's recommendations. The membranes were blocked for 1 h at room temperature in 5% bovine serum albumin (MilliporeSigma) and incubated overnight with primary anti-bodies at 4°C. After primary antibody incubation, horseradish peroxidase (HRP)-conjugated secondary antibodies were used to incubate the membranes for 2 h at room temperature. The primary antibodies used in the present study were as follows: Anti-IFN-γ (1:1,000; cat. no. ab9657), anti-TNF-α (1:1,000; cat. no. ab1793), anti-IL-6 (1:1,000; cat. no. ab208113), anti-ATP-binding cassette transporter A1 (ABCA1; 1:500; cat. no. ab18180), anti-monocyte chemoattractant protein-1 (MCP-1; 1:1,000; cat. no. ab21396), anti-fibronectin (1:1,000; cat. no. ab2413), anti-α-smooth muscle actin (α-SMA; 1:1,000; cat. no. ab5694) (all from Abcam), anti-IL-1β (1:1,000; cat. no. sc-32294), anti-COL1A2 (1:500; cat. no. sc-8788), anti-TLR4 (1:1,000; cat. no. sc-10741), anti-JAK2 (1:1,000; cat. no. sc-294) (all from Santa Cruz Biotechnology, Inc.), anti-ICAM-1 (1:1,000; cat. no. AF796), anti-VCAM-1 (1:1,000; cat. no. AF643) (both from R&D Systems, Inc.), anti-phosphorylated (P)-JAK2 (1:1,000; cat. no. 8082), anti-P-IκB (1:1,000; cat. no. 2859), anti-IκB (1:1,000; cat. no. 9242), anti-NF-κB (1:1,000; cat. no. 8242), anti-P-NF-κB (1:1,000; cat. no. 3033), anti-STAT3 (1:1,000; cat. no. 9139), anti-P-STAT3 (1:1,000; cat. no. 9145), anti-GAPDH (1:2,000; cat. no. 2118) (all from Cell Signaling Technology, Inc.) and anti-Lamin B1 (1:1,000; cat. no. 332000; Invitrogen; Thermo Fisher Scientific, Inc.). The secondary antibodies used in the present study were as follows: Anti-mouse (1:1,000; cat. no. 7076) and anti-rabbit (1:1,000; cat. no. 7074) (both from Cell Signaling Technology, Inc.). After washing, the membranes were visualized using enhanced chemiluminescence detection reagents (SuperSignal™ West Femto Maximum Sensitivity Substrate; cat. no. 34096; Thermo Fisher Scientific, Inc.) for 1 min. The signal intensity was detected using an image analyzer (ChemiDoc™XRS+; Bio-Rad Laboratories, Inc.) and was semi-quantified with Image Lab software version 5.1 (Bio-Rad Laboratories, Inc.).

Paraffin-embedded tissue sections (4 µm) were placed in a BOND-MAX Fully Automated IHC Staining System slide stainer (Leica Microsystems, Inc.) according to the following protocol. First, tissues were deparaffinized and pre-treated with the Epitope Retrieval Solution 2 (EDTA-buffer pH 8.8) at 98°C for 20 min. After washing steps, peroxidase blocking was carried out for 10 min using the BOND Polymer Refine Detection Kit (cat. no. DS9800; Leica Microsystems, Inc.). Tissues were washed again and then incubated with the primary antibodies (1:100) for 30 min at room temperature. The following primary antibodies were used: Anti-ICAM-1, anti-VCAM-1 (both from R&D Systems) and anti-monocyte + macrophage (MOMA-2; cat. no. ab33451; Abcam). The tissue sections were then incubated with post-primary for 8 min and were then incubated with polymer for 8 min and developed with DAB-Chromogen for 10 min at room temperature. Hematoxylin was used as the counterstain for 5 min at room temperature. The slides were examined using a Pannoramic MIDI slide scanner.

All data are presented as the mean ± standard error of the mean. Statistical significance was assessed using one-way analysis of variance with Tukey's multiple comparison test using GraphPad Prism 5.0 (GraphPad Software, Inc.). P<0.05 was considered to indicate a statistically significant difference.

The present study first designed STAT3, NF-κB and STAT3/NF-κB decoy ODNs (Fig. 1A). Decoy ODNs containing the DNA-binding consensus sequences of transcription factors selectively inhibit STAT3 and NF-κB by binding to the DNA-binding domains. To identify the functions of STAT3/NF-κB decoy ODNs in controlling STAT3 and NF-κB expression, EMSA was performed with nuclear extracts obtained from the aorta tissues of mice with LPS/AD-induced atherosclerosis and treated with Scr and STAT3/NF-κB decoy ODNs. Compared with in NC mice, LPS + AD + Scr atherosclerotic mice exhibited increased STAT3 and NF-κB DNA-binding activity. However, the LPS + AD + STAT/NF-κB treatment groups exhibited reduced STAT3 and NF-κB DNA-binding activity compared with the LPS + AD + Scr group (Fig. 1B). These results indicated that STAT3/NF-κB decoy ODNs effectively decreased STAT and NF-κB expression.

To investigate the effect of STAT3/NF-κB decoy ODNs on atherosclerotic mice, H&E histological analysis was performed. A marked increase in the thickness of the adventitia and media, and noticeable cellular infiltration in the adventitial layers of the aortae were detected in the LPS + AD + Scr group compared with in the NC group. Histological analysis showed that these histological changes were markedly decreased in the LPS + AD + STAT/NF-κB groups (Fig. 1C). Furthermore, no differences in body weight were determined among the groups (data not shown).

To examine the mechanism of action of STAT3/NF-κB decoy ODNs, the protein expression levels of IFN-γ, TNF-α, IL-1β and IL-6 were measured in the aorta tissues of atherosclerotic mice using western blot analysis (Fig. 1D). Notably, the expression levels of IFN-γ, TNF-α, IL-1β and IL-6 were significantly higher in the LPS + AD + Scr group than those in the NC and STAT/NF-κB groups. However, in the LPS + AD + STAT/NF-κB group, the increase in these protein expression levels was returned to a near-NC level. These observations suggested that STAT3/NF-κB decoy ODNs can inhibit atherosclerotic morphological changes and inflammation.

The tissue sections were subjected to histological analysis with Verhoeff-Van Gieson and Masson trichrome staining in order to characterize the abdominal aortic lesions. Besides cellular components, the extracellular matrix (ECM) of atherosclerotic plaque serves a relevant role in the initiation and subsequent progression of atherosclerosis. Major ECM structural and signaling components include elastin and collagen (16). Verhoeff-Van Gieson staining of elastin revealed no elastin fiber fragmentation in the NC and STAT/NF-κB groups. The elastin content in the medial layer was diminished and fragmented in the LPS + AD + Scr atherosclerosis group, perhaps because of the increased medial area or thinning of the elastin fibers; however, the elastin fibers were less disrupted in the LPS + AD + STAT/NF-κB mice compared with that in the LPS + AD + Scr mice (Fig. 2A). In addition, based on Masson's trichrome staining, a marked increase in collagen deposition in the aortae of LPS + AD + Scr group mice was detected, whereas this was markedly reduced in LPS + AD + STAT/NF-κB mice (Fig. 2C). Western blot analysis confirmed these results, showing that the protein expression levels of COL1A2 were increased in the LPS + AD + Scr group compared with those in the NC group, but were significantly decreased following treatment with STAT3/NF-κB decoy ODNs (Fig. 2B). Elastin is one of the dominant ECM proteins and collagen accumulation is characteristic of atherosclerotic plaques (16,17). These results suggested that increased ECM components in the LPS + AD + Scr mice may be attenuated by STAT3/NF-κB decoy ODNs.

To further investigate the cardioprotective effects of STAT3/NF-κB decoy ODNs, their effects on heart morphology were assessed. H&E staining of heart tissues revealed that myocardial fibers in the NC group were normal and ordered. There were no broken fibers and the nuclei of the myocardial cells were regular. In addition, no marked changes were observed between the NC and STAT/NF-κB groups. The LPS + AD + Scr mice displayed structural abnormalities, with disordered, extensively collapsed and degenerated muscle fibers; however, these histological changes were markedly alleviated by treatment with STAT3/NF-κB decoy ODNs (Fig. 3A).

ABCA1 serves a central role in the early stages of the reverse cholesterol transport pathway by mediating lipid efflux from macrophages and is important for maintaining cellular cholesterol homeostasis (18,19). Therefore, the present study explored the effects and potential mechanisms of STAT3/NF-κB decoy ODNs on ABCA1 expression in atherosclerotic mice. As shown in Fig. 3B, STAT3, NF-κB and STAT3/NF-κB decoy ODNs significantly increased the protein expression levels of ABCA1 compared with those in the LPS + AD + Scr group. Given that ABCA1 expression can be suppressed by pro-inflammatory stimuli via the NF-κB signaling pathway, STAT3/NF-κB decoy ODNs may regulate ABCA1 expression and cholesterol metabolism.

The expression of ICAM-1 and VCAM-1 in endothelial cells is the earliest known event in the initiation and progression of atherosclerosis (20). To identify the function of STAT3/NF-κB decoy ODNs in ICAM-1 and VCAM-1 expression, IHC staining was performed. IHC staining demonstrated that ICAM-1 and VCAM-1 were expressed in some endothelial cells over lesions and in the endothelia adjacent to the lesions in the LPS + AD + Scr mice, whereas the changes in ICAM-1 and VCAM-1 expression were markedly decreased in the LPS + AD + STAT/NF-κB mice (Fig. 4A and C). Western blot analysis further confirmed that ICAM-1 and VCAM-1 expression levels in the aortae of LPS + AD + Scr atherosclerotic mice were higher than those in the NC and STAT/NF-κB groups, whereas these expression levels were decreased in the LPS + AD + STAT/NF-κB mice (Fig. 4B).

To determine whether STAT3/NF-κB decoy ODNs may act on fibrosis, the expression levels of fibrosis-related proteins were detected in aortic lesions. Vascular fibrosis involves the excess accumulation of ECM proteins, such as collagen, proteoglycan and fibronectin, in the arterial wall, leading to decreased luminal diameter and increased vascular stiffness, which are characteristics of atherosclerosis (21). Western blot analysis detected the expression levels of the ECM-related marker fibronectin and the contractile smooth muscle cell (SMC)-related marker α-SMA. In the LPS + AD + Scr group the aortic protein expression levels of fibronectin and α-SMA were significantly increased compared with those in the NC and STAT/NF-κB groups. By contrast, the expression levels of fibronectin and α-SMA were significantly decreased in the LPS + AD + STAT/NF-κB mice compared with those in the LPS + AD + Scr group (Fig. 5A). MCP-1 serves a crucial role in initiating atherosclerosis by recruiting macrophages and monocytes to the vessel walls (22). The results of the present study showed that the expression levels of MCP-1 were increased in the LPS + AD + Scr group compared with those in the NC and STAT/NF-κB groups. Conversely, the expression levels of MCP-1 were inhibited in the LPS + AD + STAT/NF-κB group (Fig. 5A).

IHC staining was also performed with the common macrophage marker MOMA-2 (Fig. 5B). The progressive deposition of inflammatory cells, such as macrophages, promotes plaque vulnerability leading to enlarged unstable plaques covered by a thin fibrous cap (23). MOMA-2-positive cells were not detected in the NC and STAT/NF-κB groups, but were observed in the LPS + AD + Scr group. Moreover, macrophage accumulation was markedly reduced in the LPS + AD + STAT/NF-κB atherosclerotic lesions.

Numerous inflammatory signaling pathways that contribute to the pathogenesis of atherosclerosis are modulated by the transcription factor NF-κB, which is a master regulator of innate and adaptive immune responses (24). In addition, TLR4 serves a crucial role in the initiation of an innate immune response through activation of inflammatory cells via the NF-κB-dependent pathway (24). Therefore, the present study examined the effects of STAT3/NF-κB decoy ODNs on LPS-stimulated inflammation via the TLR4/NF-κB signaling pathway in atherosclerotic mice. Western blot analysis indicated that the protein expression levels of TLR4 were increased in the LPS + AD + Scr group compared with those in the NC group; however, were decreased following treatment with STAT3/NF-κB decoy ODNs (Fig. 5C). Because the activity of NF-κB is regulated by IκBα, the effect of STAT3/NF-κB decoy ODNs on the phosphorylation of IκBα was assessed. As shown in Fig. 5C, in a mouse model of atherosclerosis, the expression levels of P-IκBα were increased in the cytoplasm and the expression levels of P-NF-κB were increased in the nucleus. In the LPS + AD + STAT/NF-κB mice it was indicated that that STAT3/NF-κB decoy ODNs may inhibit NF-κB activation by inhibiting P-IκBα. These results suggested that STAT3/NF-κB decoy ODNs may markedly inhibit the LPS-activated TLR4/NF-κB signaling pathway in a mouse model of atherosclerosis.

The JAK/STAT signaling pathway has a vital role in various important cellular responses, such as inflammation, metabolism, cell proliferation and gene transcription (25). Western blot analysis was performed to detect the protein expression levels of JAK2 and STAT3 in the aortae. As shown in Fig. 5C, the LPS + AD + Scr group exhibited markedly enhanced P-JAK2 and P-STAT3 expression compared with that in the NC group. By contrast, the LPS + AD + STAT/NF-κB groups exhibited decreased P-JAK2 and P-STAT3 expression levels compared with those in the LPS + AD + Scr group. Taken together, these findings suggested that STAT3/NF-κB decoy ODNs may suppress inflammation in atherosclerotic mice by effectively inhibiting the STAT/NF-κB signaling pathway.