Diosgenin restores Aβ-induced axonal degeneration by reducing the expression of heat shock cognate 70 (HSC70)

Abstract

We previously found diosgenin, an herbal drug-derived steroid sapogenin, to be remarkably effective at restoring Aβ-induced axonal degeneration and improving memory function in model of Alzheimer's disease (AD), 5XFAD mouse. In this study, we investigated the downstream signaling of diosgenin and explored new therapeutic targets in AD. We showed that the expression of heat shock cognate (HSC) 70 was increased in Aβ-treated neurons and in 5XFAD mice but was decreased by diosgenin treatment. In addition, knockdown of HSC70 significantly promoted axonal growth in neurons. As an association molecule of HSC70 in neurons, α-tubulin was detected by immunoprecipitation. After Aβ treatment, α-tubulin expression was greatly reduced in the degenerated axons, suggesting that a decline in α-tubulin may be one of the factors which correlates with axonal disruption in AD pathology. We hypothesized that the degradation of α-tubulin is triggered by the chaperone activity of HSC70. However, diosgenin significantly normalized the α-tubulin level, a potentially critical process for axonal formation. Our study indicated that reducing the HSC70 level is a new possible therapeutic target of axonal regeneration in AD.

Figure 1

HSC70 expression in 5XFAD mouse cortex was drastically reduced by diosgenin administration. Diosgenin (0.1 µmol/kg/day, p.o.) or vehicle solution was administered to wild-type and 5XFAD mice (male, 5–6 months old) for 15 days. (A) On the last day of administration, an object recognition memory test was performed. The preferential indexes of the training and test session are shown. *p < 0.05 vs 5XFAD/Veh, two-tailed one-way ANOVA post hoc Dunnett’s test, ^##p < 0.01, ^###p < 0.001 vs 50%, two-tailed one sample t-test, n = 4 mice. (B) Cortex lysates were prepared, and the protein expression levels were compared on 2D-PAGE for each mouse. (C) Magnified images of the sections indicated by the red arrows in (B). This protein was more highly expressed in 5XFAD mice than in wild-type mice, but its expression was drastically decreased by diosgenin administration. This spot was cut out for MS analysis. (D) Quantitative value of the expression levels of the spot.

Figure 2

Diosgenin treatment decreased expression of HSC70 in cultured neurons and in 5XFAD mouse brains. (A,B) Mouse cortical neurons (ddY, E14) were cultured for three days and then treated with Aβ_35–25 (10 µM) or Aβ_25–35 (10 µM) for three days. Then, neurons were treated with diosgenin (0.1 or 1 µM) or vehicle solution (0.1% EtOH) for four days. Neurons were fixed and immunostained for HSC70 and GAPDH, or pNF-H. (A) The expression level of HSC70 (yellow dotted line) and GAPDH in neurons were quantified, and the expression level of HSC70 (ratio to GAPDH) were quantified in each neuron. (B) pNF-H-positive axonal lengths were quantified for each treatment group. **p < 0.01, ***p < 0.001, two-tailed one-way ANOVA post hoc Dunnett’s test. (A) n = 115–217 neurons, (B) n = 11–15 photos were quantified for these analyses. (C,D) Wild-type and 5XFAD mice (Female, 7–8 months old) were treated with diosgenin (0.1 µmol/kg/day, p.o.) or vehicle solution (sesame oil) for 18 days. (C) WB for HSC70 in the wild-type mouse and diosgenin- or vehicle-administered 5XFAD mouse cortex. Representative cropped images are shown for each group. The full-length blots are presented in Supplementary Fig. 2D. (D) Quantitative value for the expression levels of HSC70 (ratio to β-actin), n = 3–4 mice.

Figure 3

Diosgenin-induced the reduction of HSC70 and axonal regrowth are mediated by 1,25D₃-MARRS. Mouse cortical neurons were cultured for three days and then treated with or without Aβ_25–35 (10 µM) for three days. Three days after Aβ_25–35 treatment, neurons were incubated with the 1,25D₃-MARRS antibody (MARRS Ab) or normal rabbit IgG (Control Ab) for 15 minutes, followed by treatment with diosgenin (1 µM) or vehicle solution (0.1% EtOH). Four days after the treatments, the neurons were fixed and immunostained for HSC70 and GAPDH, or pNF-H. (A) The expression level of HSC70 and GAPDH in neurons were quantified, and the expression level of HSC70 (ratio to GAPDH) were quantified in each neuron. (B) pNF-H-positive axonal lengths were quantified for each treatment group. *p < 0.05, **p < 0.01, ****p < 0.0001, two-tailed one-way ANOVA post hoc Bonferroni’s multiple comparison test. (A) n = 78–243 neurons, (B) n = 9–14 photos were quantified for these analyses.

Figure 4

Effect of HSC70 knockdown on axonal growth. siRNA for HSC70 (300 µM) or control siRNA (300 µM) was transfected together with GFP vector into mouse cortical neurons. Four days later, neurons were immunostained with HSC70 and GAPDH, or pNF-H. (A) The expression level of HSC70 and GAPDH in neurons were quantified, and the expression level of HSC70 (ratio to GAPDH) were quantified in each neuron. (B) pNF-H-positive axonal lengths were quantified for each treatment group. ***p < 0.001, ****p < 0.0001, two-tailed unpaired t-test. (A) n = 87–155 neurons, (B) n = 25–28 photos were quantified for these analyses.

Figure 5

α-Tubulin was identified as a binding (client) protein of HSC70. Mouse cortical neurons were cultured for three days and then treated with vehicle solution or Aβ_25–35 (10 µM) for 30 min at 37 °C. Cell lysates were prepared and co-immunoprecipitated with the anti-HSC70 antibody. Precipitated proteins were mildly heated at 65 °C to maintain the complex formation of HSC70 and its binding partners, and electrophoresed to perform silver staining or WB for HSC70. The protein band in which the intensity was increased by Aβ treatment (red arrow) was cut out and prepared for MS analysis. The images in the figure were cropped and the full-length gels are presented in Supplementary Fig. 3.

Figure 6

Relationship between α-tubulin expression and axonal formation in Aβ- and diosgenin-treated neurons. (A,B) Mouse cortical neurons were cultured for two days and treated with Aβ_25–35 (10 µM) for one day. After the medium was removed, the neurons were treated with fresh medium for four days. Neurons were fixed and double-immunostained for α-tubulin and pNF-H or observed with DIC. (A) Structure-retained axons and dot-like or swollen structure-lost axons were observed. (B) α-Tubulin expression was quantified in the structure-retained or structure-lost axons. ^####p < 0.0001, two-tailed unpaired t-test. n = 84–104 axons were quantified for the analysis. (C–E) Mouse cortical neurons were cultured for two days and treated with Aβ_25–35 (10 µM) for one day, followed by treatment with diosgenin (1 µM) or vehicle solution (0.1% EtOH) for four days. (C) Neurons were fixed and double-immunostained for α-tubulin and pNF-H or observed with DIC. (D) The percentage of structure-lost (low expression of α-tubulin) axons is shown for each treatment group. (E) Densities of pNF-H-positive axons were quantified for each treatment. ****p < 0.0001, two-tailed one-way ANOVA post hoc Bonferroni’s multiple comparison test. (D) n = 3 photos, (E) n = 12–25 photos were quantified for these analyses.