Fig 2: DIPQUO can synergize with other known GSK3-β inhibitors to promote osteogenic differentiation. C2C12 cells were treated for 3 days with (A) subthreshold (5 μM) DIPQUO, (5 μM) CHIR, or a combination thereof or (B) subthreshold DIPQUO, (500 nM) AZD, or a combination thereof and analyzed by staining for ALP expression and by colorimetric substrate assay for ALP activity. C, ALP expression and activation normally promoted by the full dose of DIPQUO was attenuated by cotreating C2C12 cells with 10 μM XAV-939. The scale bar in panels A–C represents 200 μm. Values in ALP activity graphs are reported as the means ± SD; ∗∗∗∗p < 0.0001 in unpaired two-tailed Student's t test. Ordinary one-way ANOVA yielded F = 1868 and p < 0.0001 in panel A, F = 579.6 and p < 0.0001 in panel B, and F = 858.3 and p < 0.0001 in panel C. D, differentiating human skeletal muscle satellite cells were cotreated for 3 days with low-dose DIPQUO and CHIR (5 μM each) and then analyzed by quantitative RT-PCR for expression of transcripts associated with osteoblast specification, compared with treatment alone. Relative expression levels compared with DMSO-treated samples and normalized to glyceraldehyde 3-phosphate dehydrogenase are reported as the means ± SD; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001 in unpaired two-tailed Student's t test. Ordinary one-way ANOVA yielded F = 29.08 and p = 0.0008 for hDLX5 (DISTAL-LESS HOMEOBOX 5); F = 9.654 and p = 0.0133 for hOSX (OSTERIX); F = 81.15 and p < 0.0001 for hOPN (OSTEOPONTIN); and F = 110.1 and p < 0.0001 for hGPNMB (OSTEOACTIVIN). E, synergistic activation of p38 MAPK was analyzed by Western blotting after 2.5 h treatment with subthreshold concentrations of AZD (100 nM) and DIPQUO (2.5 μM), compared with cotreatment using the same concentrations. Relative phospho-p38 MAPK levels are reported as the means ± SD; ∗p < 0.05. Ordinary one-way ANOVA yielded F = 3.521 and p = 0.0686. F, C2C12 cells were either mock-transfected or transfected for 24 h with hemagglutinin-tagged GSK3-β containing a serine-to-alanine mutation (S9A), treated with DIPQUO or DMSO vehicle, and analyzed by Western blotting for construct expression. G, equivalent samples were analyzed for ALP activity using colorimetric assay for pNPP substrate after 3 days. Activity levels are normalized to the total protein and reported as the mean ± SD; ∗p < 0.05 and ∗∗p < 0.01 in unpaired two-tailed Student's t test. Ordinary one-way ANOVA yielded F = 25.60 and p = 0.0002. In all figure panels, representative images are shown from at least three biological replicates. ALP, alkaline phosphatase; AZD, AstraZeneca GSK3-β-specific inhibitor AZD2858; CHIR, Wnt signaling activator CHIR99021; GSK3-β, glycogen synthase kinase 3-beta; pNPP, p-nitrophenylphosphate.

Fig 3: 12B2 and 15C2 are specific for nonphospho-Ser GSK3ß/a peptides. Each antibody was screened in indirect ELISA titers against npS9 GSK3ß, pS9 GSK3ß, npS21 GSK3a and pS21 GSK3a peptides (n = 3 independent experiments). (A) 12B2 showed strong reactivity for npS9 GSK3ß compared to npS21 GSK3a peptides and did not react with pS9 or pS21 GSK3 peptides (EC50 values: npS9 = 2.1 nM; pS9 = indeterminate (id); npS21 = 6.4 nM; pS21 = id). (B) To further confirm the specificity of 12B2, ELISAs were performed by coating wells with a wide range of peptide amounts (0 – 6.4 µg peptide/well). 12B2 showed strong reactivity with the npS GSK3 peptides (ß > a), but did not react with pS GSK3 peptides. (C) 15C2 showed stronger reactivity for npS21 GSK3a compared to npS9 GSK3ß and did not react with pS9 or pS21 GSK3 peptides (EC50 values: npS9 = 2.4 nM; pS9 = id; npS21 = 277 pM; pS21 = id). (D) To further confirm the specificity of 15C2, ELISAs were performed by coating wells with a wide range of peptide amounts (0 – 6.4 µg peptide/well). 15C2 showed strong reactivity with the npS GSK3 peptides, but did not react with pS GSK peptides. It is noteworthy that synthetic peptides provide a homogeneous source of modified peptides, and thus, are ideal for challenging the specificity of the antibodies against nonphospho-Ser and phospho-Ser residues in GSK3.

Fig 4: Effects of quercetin on MAPK, AMPKα and GSK3-α/β signals in human tongue squamous carcinoma SAS cells. Cells were treated with quercetin (50 µM) for 30–120 min. The protein expression levels of p-ERK1/2, p-JNK1/2, p-p38, p-AMPKα and p-GSK3-α/β were determined by western blot analysis. Images are representative of at least three independent experiments. Semi-quantification was determined by densitometric analysis. Values are expressed as the mean ± standard deviation of three independent experiments. *P<0.05 vs. vehicle control. AMPK, adenosine monophosphate-activated protein kinase; GSK, glycogen synthase kinase; p-, phosphorylated.

Fig 5: Fluorescent gel shift-based inhibition assay of GSK3 activity as a function of inhibitor (lithium chloride, LiCl) concentration: (A) Fluorescent gel image from the GSK3 inhibition assay using LiCl (0−1.25 mM) in the presence of T-Pep(p). T-Pep(p) was reacted with recombinant active GSK3 at varied concentrations of LiCl: 0 μM (Lane 1), 200 μM (Lane 2), 500 μM (Lane 3), 750 μM (Lane 4), 1 mM (Lane 5), and 1.25 mM (Lane 6); (B) Inhibitory graph of GSK3 activity as a function of LiCl concentration. GSK3 activity (%) was calculated and plotted from the relative fluorescence intensities in the two band regions which corresponded to T-Pep(p) and T-Pep(pp) on agarose, according to Equation (1). The final concentration of GSK3 was 1.25 U in 20 μL of reaction volume. The standard deviation was obtained from three independent experiments.