The decoy oligodeoxynucleotide against HIF-1α and STAT5 ameliorates atopic dermatitis-like mouse model

Abstract

Atopic dermatitis (AD) is a common inflammatory skin disease caused by an immune disorder. Mast cells are known to be activated and granulated to maintain an allergic reaction, including rhinitis, asthma, and AD. Although hypoxia-inducible factor-1 alpha (HIF-1α) and signal transducer and activator of transcription 5 (STAT5) play crucial roles in mast cell survival and granulation, their effects need to be clarified in allergic disorders. Thus, we designed decoy oligodeoxynucleotide (ODN) synthetic DNA, without open ends, containing complementary sequences for HIF-1α and STAT5 to suppress the transcriptional activities of HIF-1α and STAT5. In this study, we demonstrated the effects of HIF-1α/STAT5 ODN using AD-like in vivo and in vitro models. The HIF-1α/STAT5 decoy ODN significantly alleviated cutaneous symptoms similar to AD, including morphology changes, immune cell infiltration, skin barrier dysfunction, and inflammatory response. In the AD model, it also inhibited mast cell infiltration and degranulation in skin tissue. These results suggest that the HIF-1α/STAT5 decoy ODN ameliorates the AD-like disorder and immunoglobulin E (IgE)-induced mast cell activation by disrupting HIF-1α/STAT5 signaling pathways. Taken together, these findings suggest the possibility of HIF-1α/STAT5 as therapeutic targets and their decoy ODN as a potential therapeutic tool for AD.

Keywords: HIF-1α; MT: Oligonucleotides: Therapies and Applications; STAT5; atopic dermatitis; decoy oligodeoxynucleotide; mast cell.

Graphical abstract

Figure 1

Structure of the synthetic HIF-1α/STAT5 decoy ODN and transfection efficiency of the HIF-1α/STAT5 decoy ODN in an AD-like in vitro and in vivo model (A) Primary sequence of the HIF-1α/STAT5 decoy ODN. (B) The fluorescence result showed that HIF-1α/STAT5 decoy ODN was effectively transfected into RBL-2H3 cells. The RBL-2H3 cells were transfected with FITC-labeled HIF-1α/STAT5 decoy ODN (60 nM; green). The cells were then stained with DAPI (blue). Scale bar, 50 μm. (C) Transfection efficiency of the FITC-labeled HIF-1α/STAT5 decoy ODN was analyzed using flow cytometry. (D and E) HIF-1α- and STAT5 DNA-binding activity was analyzed using electrophoretic mobility shift assay (EMSA) using the nuclear extract of the RBL-2H3 cells. Three independent EMSA data were used to quantify the dot plots. +, treated; −, un-treated; vehicle (Veh), distilled water; scrambled (Scr) ODN, scrambled decoy ODN; HIF-1α/STAT5 ODN, HIF-1α/STAT5 decoy ODN; NT, IgE+Ag non-treated; IgE+Ag, IgE+Ag treated. ∗p < 0.05 compared with the vehicle group; ^†p < 0.05 compared with the IgE+Ag-sensitized group. (F and G) DNA-binding activity of HIF-1α and STAT5 was measured using EMSA using the mouse nuclear extract. The dot graphs were quantified from three independent EMSA data. +, treated; −, un-treated; vehicle (Veh), distilled water; scrambled (Scr) ODN, scrambled decoy ODN; HIF-1α/STAT5 ODN, HIF-1α/STAT5 decoy ODN; NT, DNCB and DfE non-treated group; DNCB/DfE, DNCB- and DfE-sensitized group. ∗p < 0.05 compared with the vehicle group; ^†p < 0.05 compared with the DNCB/DfE-sensitized group.

Figure 2

The HIF-1α/STAT5 decoy ODN significantly suppressed the expression of HIF-1α and STAT5 and attenuated their downstream target genes in IgE+Ag-sensitized RBL-2H3 cells (A) Representative images of the immunofluorescence staining for HIF-1α (green, top) and STAT5 (red, bottom) in RBL-2H3 cells. The nuclei were labeled with DAPI (blue). Scale bar, 50 μm. (B) Quantification of the HIF-1α and STAT5 immunofluorescence signals. (C) Western blot analysis of HIF-1α and HIF-1α-related genes, including VEGF, iNOS, and COX-2. GAPDH was used as a loading control. (D–G) The dot graphs were quantified from three independent immunoblot data. (H) Immunoblot analysis of p-STAT5, t-STAT5, Bcl2, Bcl-_XL, and cyclin D3. GAPDH was used as a loading control. (I–L) Three independent immunoblots were used to quantify the dot graphs. +, treated; −, un-treated; vehicle (Veh), distilled water; scrambled (Scr) ODN, scrambled decoy ODN; HIF-1α/STAT5 ODN, HIF-1α/STAT5 decoy ODN; NT, IgE+Ag non-treated; IgE+Ag, IgE+Ag treated. ∗p < 0.05 compared with the vehicle group; ^†p < 0.05 compared with the IgE+Ag-sensitized group.

Figure 3

Effect of the HIF-1α/STAT5 decoy ODN on IgE-mediated degranulation in IgE+Ag-challenged RBL-2H3 cells (A) Immunofluorescence staining for tryptase (green). The nuclei were labeled with DAPI (blue). Scale bar, 20 μm. (B) Quantification of the tryptase fluorescence signal. The dot graphs were quantified from three independent immunofluorescence images. (C) Immunoblot result showed that the HIF-1α/STAT5 decoy ODN attenuated the cytokine expression. GAPDH was used as a loading control. (D–F) The graphs show the quantification of (D) tryptase, (E) TNF-α, and (F) IL-4, each normalized to GAPDH. The graphs were quantified from three independent immunoblot data. +, treated; −, un-treated; vehicle (Veh), distilled water; scrambled (Scr) ODN, scrambled decoy ODN; HIF-1α/STAT5 ODN, HIF-1α/STAT5 decoy ODN; NT, IgE+Ag non-treated; IgE+Ag, IgE+Ag treated. ∗p < 0.05 compared with the vehicle group; ^†p < 0.05 compared with the IgE+Ag-sensitized group.

Figure 4

The HIF-1α/STAT5 decoy ODN-induced apoptosis of RBL-2H3 cells via the regulation of apoptosis-associated molecules (A) TUNEL staining images of the RBL-2H3 cells from the indicated groups. DAPI was used to stain nuclei (blue). After TUNEL staining, these representative images were taken from random sites locations. Apoptotic cells are indicated by arrows, and representative images of apoptotic cells are indicated in the enlarged frames. Scale bar, 20 μm. (B) Immunoblotting analysis of pro-apoptotic proteins. GAPDH was used as a loading control. (C–E) The graphs show the quantification of (C) cytochrome c, (D) cleaved caspase-3, and (E) Bax, each normalized to GAPDH. The graphs were quantified from three independent immunoblot data. (F) Western blot analysis of anti-apoptotic proteins. All samples were loaded in equal volumes, as normalized by loading GAPDH together. (G and H) The graphs show the quantification of (G) Bcl-2 and (H) Bcl-_XL, each normalized to GAPDH. The dot graphs were quantified from three independent immunoblot data. +, treated; −, un-treated; vehicle (Veh), distilled water; scrambled (Scr) ODN, scrambled decoy ODN; HIF-1α/STAT5 ODN, HIF-1α/STAT5 decoy ODN; NT, IgE+Ag non-treated; IgE+Ag, IgE+Ag treated. ∗p < 0.05 compared with the vehicle group; ^†p < 0.05 compared with the IgE+Ag-sensitized group.

Figure 5

Effect of the HIF-1α/STAT5 decoy ODN in DNCB/DfE-induced AD-like symptoms, skin histological changes, and mast cell infiltration (A) The schedule for induction of AD caused by DNCB/DfE sensitization and treatment of HIF-1α/STAT5 decoy ODN. (B) Balb/c mice dorsal skin lesions of each group. The animal experiment was performed in two independent experiments (n = 4 in the first experiment and n = 5 in the second experiment, for a total of n = 9 mice per group). (C) Paraffin-embedded sections of murine dorsal skin stained with hematoxylin and eosin. Scale bar, 100 μm. (D and E) The thicknesses of the (D) epidermis and (E) dermis were measured from at least 10 random fields per section at 200× magnification. (F) Paraffin-embedded sections of murine dorsal skin stained with Giemsa (n = 9). The representative images of mast cells are indicated in the enlarged frames. Scale bar, 50 μm. (G–I) The graphs show the statistical analysis of (G) the number of infiltrated mast cells, (H) the number of degranulated mast cells, and (I) degranulated mast cell rate (degranulated/non-degranulated). The number of infiltrated or degranulated in mast cells was counted from at least 10 random fields per skin section at 400× magnification. (J) Serum IgE was measured using mouse serum (n = 5). Vehicle (Veh), distilled water; scrambled (Scr) ODN, scrambled decoy ODN; HIF-1α/STAT5 ODN, HIF-1α/STAT5 decoy ODN; NT, DNCB and DfE non-treated group; DNCB/DfE, DNCB- and DfE-sensitized group. ∗p < 0.05 compared with the vehicle group; ^†p < 0.05 compared with the DNCB/DfE-sensitized with Scr ODN group.

Figure 6

The HIF-1α/STAT5 decoy ODN improved the DNCB/DfE-induced Th2 inflammation and skin barrier destruction (A) The representative tryptase in immunohistochemistry (n = 9). Scale bar, 100 μm. (B) The graphs show the tryptase-positive area (%). (C) Tryptase protein expression levels were detected using immunoblotting. GAPDH was used as a loading control. (D) The graphs show the quantification of tryptase. The dot graphs were quantified from three independent immunoblot data. (E) The protein expression levels of pro-inflammatory and Th2 cytokines were detected using immunoblotting with the indicated antibodies. GAPDH was used as a loading control. (F–I) The graphs show the quantification of (F) TNF-α, (G) IL-1β, (H) IL-4, and (I) TSLP, each normalized to GAPDH. The dot graphs were quantified from three independent immunoblot data. (J) Paraffin-embedded skin sections were deparaffinized and stained with anti-filaggrin (green) (n = 9), and the nuclei were stained with DAPI (blue). The bottom image is an enlargement of the area marked with a red square. The scale bars of the upper images are 100 μm, whereas those of the bottom images are 50 μm. (K) The immunoblotting analysis shows the expression levels of filaggrin and GAPDH in the mouse skin tissue. (L) The bar graphs show the quantitative signal intensity of filaggrin after normalization with GAPDH. The bar graphs were quantified from three independent immunoblot data. +, treated; −, un-treated; vehicle (Veh), distilled water; scrambled (Scr) ODN, scrambled decoy ODN; HIF-1α/STAT5 ODN, HIF-1α/STAT5 decoy ODN; NT, DNCB and DfE non-treated group; DNCB/DfE, DNCB- and DfE-sensitized group. ∗p < 0.05 compared with the vehicle group; ^†p < 0.05 compared with the DNCB/DfE-sensitized group.

Figure 7

Inhibitory effect of the HIF-1α/STAT5 decoy ODN on HIF-1α and STAT5 expression levels and their downstream target genes in DNCB/DfE-induced AD-like skin disorder (A) Representative images of HIF-1α (green, top) and STAT5 (red, bottom) in the mouse dorsal skin via immunofluorescence staining (n = 9). The nuclei were stained with DAPI (blue). Scale bar, 200 μm. (B) Quantification of the immunofluorescence signals of HIF-1α and STAT5. (C) Western blot analysis of HIF-1α and HIF-1α-related genes, including VEGF, iNOS, and COX-2. GAPDH was used as a loading control. (D–G) The dot graphs show the quantitative signal intensity of (D) HIF-1α, (E) VEGF, (F) iNOS, and (G) COX-2 after normalization with GAPDH. The graphs were quantified from three immunoblot data. (H) The protein expression levels of STAT5 and STAT5-targeted genes, including Bcl2, Bcl-_XL, and cyclin D3, were analyzed using immunoblotting. GAPDH was used as a loading control. (I–L) The bar graphs show the quantitative signal intensity of (I) STAT5, (J) Bcl2, (K) Bcl-_XL, and (L) cyclin D3 after normalization with GAPDH or t-STAT5. The dot graphs were quantified from three independent immunoblot data. (M) The graphic illustration of the effect of HIF-1α/STAT5 decoy ODN in DNCB/DfE-induced and IgE/Ag-induced AD-like model via suppressing of transcriptional expression. +, treated; −, un-treated; vehicle (Veh), distilled water; scrambled (Scr) ODN, scrambled decoy ODN; HIF-1α/STAT5 ODN, HIF-1α/STAT5 decoy ODN; NT, DNCB and DfE non-treated group; DNCB/DfE, DNCB- and DfE-sensitized group. ∗p < 0.05 compared with the vehicle group; ^†p < 0.05 compared with the DNCB/DfE-sensitized group.