(A) Chemical structures of deoxygedunin and two other derivatives. (B) Gedunin derivatives suppress glutamate-provoked neuronal cell death. Hippocampal neurons were pretreated with 0.5 µM gedunins for 30 min, followed by 50 µM glutamate treatment for 16 h. Quantitative analysis of neuronal apoptosis was conducted. (C) Deoxygedunin inhibits OGD-triggered neuronal apoptosis in a dose-dependent manner.

(A) Deoxygedunin triggers TrkB activation in BDNF conditional knockout cortex. 2–3 months old BDNF cortex conditional knockout mice were intraperitoneally injected with 5 mg/kg deoxygedunin. In 4 h, the mice were sacrificed and brain lysates were analyzed by immunoblotting. (B & C) Deoxygedunin and 7,8-DHF prevent vestibular ganglion loss in BDNF −/− pups. BDNF +/− mice were mated with the same genotypes of mice. At E7.5 days, the pregnant mothers were administrated with 5 mg/kg deoxygedunin or 7,8-DHF until birth. The neonatal pups continued on drug treatment for 1 or 2 days till death. The inner ear sections were stained with toluidine blue.

(A) Forced swim test. 2–3-month-old Male mice (n = 8 mice/group) were treated with imipramine (20 mg/kg), deoxygedunin (5 mg/kg), 7,8-DHF (5 mg/kg) and vehicle solvent saline by intraperitoneal injection for 5 days, and subjected to a forced swim test (6 min, immobility recorded in the last 4 min). (B) Forced swim test with TrkB F616A knockin mice. Male TrkB knockin mice were given the regular drinking water or 1NMPP1 containing drinking water one day before the drugs. Deoxygedunin/1NMPP1 group is significantly different from the water group but not from control in 1NMPP1. Data are presented as mean ± SEM of n = 8–10 mice/group (**P<0.01 and ***P<0.001 against vehicle control group, One-way ANOVA, Dunnett's test).

(A) [³H]deoxygedunin binds the ECD but not ICD of TrkB receptor. In vitro binding assay with purified TrkB ECD or ICD (10 µg) and [³H]deoxygedunin (upper panel). In vitro binding assay with purified fragments of TrkB ECD (10 µg) and [³H]deoxygedunin. (B) Determination of binding constant. Binding curve (upper panel). Scatchard plot analysis revealed that deoxygedunin binds TrkB with binding constant of 1.4 µM (lower panel). (C) Deoxygedunin provokes TrkB dimerization. mGST-TrkB and HA-TrkB were cotransfected into HEK293 cells, treated with 0.5 µM various gedunins for 30 min. GST-TrkB was pulled down by glutathione beads, the coprecipitated proteins were analyzed with anti-HA HRP antibody. (D) Deoxygedunin induces TrkB autophosphorylation. HEK 293 cells were transfected with various Trk receptors, followed by deoxygedunin treatment for 30 min. Cell lysates were analyzed by immunoblotting assay.

(A) Deoxygedunin prevents glutamate-triggered neuronal apoptosis in wild-type but not TrkB null neurons. Cortical neurons were prepared from the P0 pups (TrkB +/− x TrkB +/−). The neurons were pretreated with a variety of compounds as indicated for 30 min, followed by 50 µM glutamate for 16 h. The cell lysates were analyzed by immunoblotting with anti-p-TrkB, anti-p-TrkA and anti-active caspase-3 etc. (B) Deoxygedunin diminishes caspase-3 activation regardless of TrkC genotype. (C) Deoxygedunin selectively activates TrkB F616A, which can be blocked by 1NMPP1. Primary cortical neurons were prepared from TrkB F616A knockin mice. The primary cultures were pretreated for 30 min with either K252a Trk tyrosine kinase inhibitor (100 nM) or 1NMPP1 inhibitor (100 nM), followed by 0.5 µM Deoxygedunin for 30 min. Immunoblotting with various antibodies were performed. (D) Deoxygedunin suppresses KA-induced neuronal cell death in TrkB F616A mutant mice, which can be blocked by 1NMPP1. TrkB F616A knockin mice were treated with following reagents: saline, 1NMPP1, kainic acid, 1NMPP1 + kainic acid, deoxygedunin, 1NMPP1 + deoxygedunin, 1NMPP1 + deoxygedunin + kainic acid, as described in experimental section. Immunoblotting was conducted with indicated antibodies.

(A) Deoxygedunin activates TrkB in primary hippocampal neurons. Hippocampal neurons were treated with 500 nM gedunin derivatives for 30 min and neurons were fixed and immunostained with rabbit polyclonal anti-p-TrkB (816) (1∶100) and anti-MAP2. The nuclei were stained with DAPI. BDNF and a few gedunin derivatives selectively triggered TrkB phosphorylation in neurons. (B) Deoxygedunin triggers TrkB activation in primary neurons. Rat cortical neurons were treated with various concentrations of deoxygedunin for 30 min. Neuronal lysates were subjected to immunoblotting analysis by mouse monoclonal anti-p-TrkB (817)(1∶20,000). Equal amount of TrkB was loaded (anti-TrkB from Biovision, 1∶1,000) (lower panel). (C) Deoxygedunin provokes Akt and Erk1/2 activation in primary neurons in a dose and time-dependent manner. Mouse monoclonal anti-TrkB 817 was used at 1∶20,000 dilution. (D) K252a blocks deoxygedunin's agonistic effect on TrkB. Cortical neurons were pretreated with K252a (100 nM) for 30 min, followed by BDNF (100 ng/ml) or deoxygedunin (500 nM) for 30 min. Cell lysates were analyzed by immunoblotting. (E) Deoxygedunin activates TrkB in mouse brain. Deoxygedunin (5 mg/kg) was intraperitoneally injected into mice and mouse brains were dislocated at different time points. Brain lysates were analyzed by immunoblotting.

(A) Outline of Deoxygedunin (DG) fear conditioning experiment. Mice were handled and habituated to testing context on the first two days, followed by systemic DG (5 mg/kg, i.p.) 1 hr prior to fear conditioning (5, 0.5mA shock-tone pairings), followed by testing in the absence of drug on days 4–5. (B) Shock reactivity during fear conditioning, demonstrating that the acute effects of the drug did not affect pain or fear responsiveness. (C) Total freezing to the conditioned cue was significantly greater on both the first and second testing day in the mice that received DG with fear conditioning. (D & E), Freezing activity during the habituation period and first 4 CS presentations on test day 1 (D) or 2 (E). There was no difference in animal activity in the test chamber prior to the onset of the conditioned cue, or between cue presentations. However there was significantly increased fear, as measured with conditioned freezing, during conditioned stimuli (CS). Dark bars represent tone conditioned stimulus presentations; **, p<0.01, *, p<0.05 between DG and vehicle groups.