Progranulin improves neural development via the PI3K/Akt/GSK-3β pathway in the cerebellum of a VPA-induced rat model of ASD

Abstract

Autism spectrum disorder (ASD) is a neurodevelopmental disease featuring social interaction deficits and repetitive/stereotyped behaviours; the prevalence of this disorder has continuously increased. Progranulin (PGRN) is a neurotrophic factor that promotes neuronal survival and differentiation. However, there have not been sufficient studies investigating its effect in animal models of autism. This study investigated the effects of PGRN on autistic phenotypes in rats treated with valproic acid (VPA) and assessed the underlying molecular mechanisms. PGRN was significantly downregulated in the cerebellum at postnatal day 14 (PND14) and PND35 in VPA-exposed rats, which simultaneously showed defective social preference, increased repetitive behaviours, and uncoordinated movements. When human recombinant PGRN (r-PGRN) was injected into the cerebellum of newborn ASD model rats (PND10 and PND17), some of the behavioural defects were alleviated. r-PGRN supplementation also reduced cerebellar neuronal apoptosis and rescued synapse formation in ASD rats. Mechanistically, we confirmed that PGRN protects neurodevelopment via the PI3K/Akt/GSK-3β pathway in the cerebellum of a rat ASD model. Moreover, we found that prosaposin (PSAP) promoted the internalisation and neurotrophic activity of PGRN. These results experimentally demonstrate the therapeutic effects of PGRN on a rat model of ASD for the first time and provide a novel therapeutic strategy for autism.

Fig. 1. Prenatal exposure to VPA caused the alterations of PGRN temporal expression in the cerebellum.

Representative blots (A) and quantification (B) showed the expression of PGRN at four different time points. Representative images (bar = 100 μm) (C) and quantification (D) of immunofluorescence staining described that the expression of PGRN is in the same trend as Western blotting. E PGRN ELISA protein quantitation of cerebellar tissue. Data are expressed as mean ± SEM. Two-way ANOVA followed by Sidak post hoc. (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, sample sizes (n): n = 5/group for Western blotting and immunofluorescence staining. n = 4/group for ELISA).

Fig. 2. Prenatal exposure to VPA induced the defection of behaviours.

Cerebellum-related behaviours: righting reflex (A), static beams (B), gait analysis (C), traction test (D). Autism-like behaviours: open field test (E), self-grooming test (F), juvenile social play (G), the three-chamber social test (H). Data are expressed as mean ± SEM. Student’s t-test (A, C_{(Fore, Hind, Gait)}, D_{(Traction score)}, E_(Faeces), G_(Social), H_{(10 min: Stranger1, A chamber, B chamber)}, H_{(20 min: A chamber, B chamber)}); Mann–Whitney test (B_{(Time to Complete, Footfalls per Trial)}, D(_{Hanging time)}, G_{(Dig, Attack)}, H_{(10 min: Centre chamber)}, H_{(20 min: Stranger1, Stranger2)}); Welch’s correction (B_{(Proportion of Trials)}, E_{(Centre grid, All grid, Climbing)}, F, H_{(10 min: Object)}, H_{(20 min: Centre chamber)}). (ns no significant difference, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, Sample sizes (n): n (CON) = 13, n (VPA) = 11).

Fig. 3. Dynamic changes of neuronal apoptosis and synaptic dysplasia in the cerebellum of VPA-induced rats.

A, B Immunoblots and quantification exhibited expression levels of Caspase3, Bcl2, and Bax at different neurodevelopmental stages. C, D Representative images (bar = 50 μm) and quantification of TUNEL staining also showed increased neuronal apoptosis in the VPA group. E, F Western blotting showed synaptic markers, PSD95 and SYP, in the cerebellum. G, H Immunofluorescence staining (bar = 25 μm) exhibited consistent alterations in PSD95 and SYP protein expressions in VPA-induced rats. Data are expressed as mean ± SEM. Two-way ANOVA followed by Sidak post hoc. (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, sample sizes (n): n = 5/group).

Fig. 4. r-PGRN improved the defection of behaviours, and PI3K inhibitor wortmannin blocked this effect in VPA-induced rats.

Cerebellum-related behaviours: righting reflex (A), static beams (B), gait analysis (C), traction test (D). Autism-like behaviours: open field test (E), self-grooming test (F), juvenile social play (G), and the three-chamber social test (H). Data are expressed as mean ± SEM. One-way ANOVA followed by Tukey’s post hoc (B_{(Proportion of Trials)}, C_{(Fore, Hind, Gait)}, D_{(Traction score)}, E_{(All grid, Faeces)}, F, G_(Social), H_{(10 min: Stranger1, A chamber, B chamber)}, H_{(20 min: A chamber, B chamber)}); post hoc Kruskal–Wallis test for multiple comparisons (A, B_{(Time to Complete, Footfalls per Trial)}, D_{(Hanging time)}, E_{(Centre grid, Climbing)}, G_{(Dig, Attack)}, H_{(10 min: Object, Centre chamber)}, H_{(20 min: Stranger1, Stranger2, Centre chamber)}). (ns no significant difference, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, sample sizes (n): n (CON) = 13, n (VPA + Vehicle) = 11, n (r-PGRN) = 11, n (r-PGRN + Wort) = 11).

Fig. 5. r-PGRN reduced neuronal apoptosis and improved synaptic development, while PI3K inhibitor wortmannin reversed this effect in VPA-induced rats.

Representative blots (A) and quantification (B) of Caspase-3, BCL-2, and Bax in the four groups. C, D Representative images (bar = 50 μm) and quantification of TUNEL staining. Representative blots (E) and quantification (F) of synaptic markers such as PSD95 and SYP. G, H Immunofluorescence staining (bar = 25 μm) exhibited consistent changes in PSD95 and SYP protein expressions. Representative images of Purkinje cell dendrites (I) that were used for quantification of spines (bar = 10 μm) and quantification (J) of the number of spines. Data are expressed as mean ± SEM. One-way ANOVA followed by Tukey’s post hoc (B_{(Caspase-3, Bax)}, D, F_(SYP), H); post hoc Kruskal–Wallis test for multiple comparisons (B_(Bcl2), F_(PSD95), J). (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, sample sizes (n): n = 5/group).

Fig. 6. PGRN activated PI3K/Akt/GSK-3β signalling pathway and the expression of M6PR, SORT1 and PSAP in VPA-induced rats at PND35.

Representative Western blots (A, C) and densitometric quantification (B, D) of p-Akt (Ser473), Akt, p-GSK-3β (Ser9) and GSK-3β expressions in the cerebellum. Representative Western blots (E) and densitometric quantification (F–H) of M6PR, SORT1 and PSAP. Data are expressed as mean ± SEM. One-way ANOVA followed by Tukey’s post hoc (B, D); Student’s t test (F, H); Mann–Whitney test (G). (*P < 0.05, **P < 0.01, ***P < 0.001, sample sizes (n): n = 5/group).