Antipurinergic therapy corrects the autism-like features in the poly(IC) mouse model

Abstract

Background: Autism spectrum disorders (ASDs) are caused by both genetic and environmental factors. Mitochondria act to connect genes and environment by regulating gene-encoded metabolic networks according to changes in the chemistry of the cell and its environment. Mitochondrial ATP and other metabolites are mitokines-signaling molecules made in mitochondria-that undergo regulated release from cells to communicate cellular health and danger to neighboring cells via purinergic signaling. The role of purinergic signaling has not yet been explored in autism spectrum disorders.

Objectives and methods: We used the maternal immune activation (MIA) mouse model of gestational poly(IC) exposure and treatment with the non-selective purinergic antagonist suramin to test the role of purinergic signaling in C57BL/6J mice.

Results: We found that antipurinergic therapy (APT) corrected 16 multisystem abnormalities that defined the ASD-like phenotype in this model. These included correction of the core social deficits and sensorimotor coordination abnormalities, prevention of cerebellar Purkinje cell loss, correction of the ultrastructural synaptic dysmorphology, and correction of the hypothermia, metabolic, mitochondrial, P2Y2 and P2X7 purinergic receptor expression, and ERK1/2 and CAMKII signal transduction abnormalities.

Conclusions: Hyperpurinergia is a fundamental and treatable feature of the multisystem abnormalities in the poly(IC) mouse model of autism spectrum disorders. Antipurinergic therapy provides a new tool for refining current concepts of pathogenesis in autism and related spectrum disorders, and represents a fresh path forward for new drug development.

Figure 1. Correction of the Core Behavioral Features of the Maternal Immune Activation (MIA) Mouse Model of Autism Spectrum Disorders (ASD).

(A) Social Preference. MIA males had a 54% decrease in social preference compared to controls (PIC-Sal 12.5+/−4.2% vs Sal-Sal 27.6+/−2.6%; one-way ANOVA F(3,42) = 3.74; with Newman-Keuls post-hoc testing; n = 9–15 males per group; age = 10-weeks; p<0.02). This was corrected by suramin treatment (PIC-Sur 28.1% vs Sal-Sal 27.6%; p = ns). (B) Social Preference as the time spent with stranger mouse vs. inanimate cup from 0–5 minutes. Analyzed by 2-Way ANOVA with Bonferroni pair-wise post testing (*p<0.05; ***p<0.001; ****p<0.0001). Treatment with suramin had little effect on normal behavior (Sal-Sal vs Sal-Sur), but a strong effect in improving social behavior in the MIA group (PIC-Sal vs. PIC-Sur). Zone x treatment interaction F(3,43) = 3.72; p<0.05; n = 9–15 males per group; age = 10-weeks. (C) Rotarod Training Curves. MIA (PIC-Sal) animals displayed deficits that were corrected by suramin treatment. Analyzed by repeated measures ANOVA with Tukey post testing: Sal-Sal vs. PIC-Sal q = 6.749, p<0.01; PIC-Sal vs PIC-Sur q = 11.13, p<0.001; n = 9–16 males per group; age = 11-weeks. (D) Rotarod Sensorimotor Coordination. MIA animals had a 28% decrease in sensorimotor coordination as measured by latency to fall by rotarod testing (PIC-Sal = 17.7+/−1.6 sec vs Sal-Sal = 24.5+/−2.2 sec; one-way ANOVA F(3,46) = 7.08; n = 9–16 males per group; age = 11-weeks; p<0.001). This was corrected by suramin treatment (PIC-Sur 27.2+/−1.6 sec vs Sal-Sal 24.5+/−2.2 sec; p = ns). Values are expressed as mean +/− SEM.

Figure 2. Relative Hypothermia Was Corrected, and Aerobic Metabolism was Increased by Antipurinergic Therapy.

(A) Relative Hypothermia in the MIA Model was Corrected by Antipurinergic Therapy. (Linear mixed effects model analysis; F(1,47) = 25.3; n = 9–16 males per group; ages 8–16 weeks; p<0.001) (B) Correction of the Relative Hypothermia Was Lost After Discontinuing Antipurinergic Therapy. Weekly injections of suramin were discontinued in females at 18 weeks of age (PIC-SUR group; orange line, inverted triangles). By 22 weeks, hypothermia in the MIA animals returned to the untreated level approximately 0.5° below normal. (F(1,39) = 43.7; n = 9–16 females per group; p<0.001). (C) Relative Hypothermia is a Long-term Feature of the Poly(IC) MIA Model. Hypothermia persisted for at least 8 months of age (linear mixed effects model analysis F(1,19) = 114; n = 9–12 females per group; p<0.001). (D) Aerobic Metabolism. Oxygen consumption in the MIA animals showed a trend toward being decreased in both sleep (light) and active (dark) cycles. Suramin treatment increased sleep cycle oxygen consumption by 11%; MIA = PIC-Sal VO₂ = 3552+/−47.6 ml/kg/hour; Treated MIA = PIC-Sur = 3938+/−45.9 (one-way ANOVA F(3,44) = 8.0; n = 6 males per group; age = 14 weeks; p = 0.0002). Antipurinergic therapy had no significant effect on oxygen consumption in the control animals; Saline-treated Controls = Sal-Sal VO₂ = 3652+/−72.8; Treated Controls = Sal-Sur = 3821+/−71.5 (n = 6 males per group; p = 0.11). Values are expressed as mean +/− SEM.

Figure 3. Plasma Immunoglobulins and Corticosterone.

(A) Plasma immunoglobulins were increased 18% by antipurinergic therapy. (PIC-Sal = 1.6+/−0.05 mg/dl; PIC-Sur = 1.9+/−0.07 mg/dl; n = 6–10 males per group; age = 12 weeks; one-way ANOVA F(3,37) = 5.72; p<0.05) (B) Plasma corticosterone levels were increased 50% by weekly suramin treatment (PIC-Sal = 77+/−14 ng/ml; PIC-Sur = 117+/−16 ng/ml; two-way ANOVA F(1,37) = 5.16; p = 0.03; n = 8–12 males per group; age = 12 weeks). Values are expressed as mean +/− SEM.

Figure 4. Cerebral Synaptosomal Ultrastructural Abnormalities Were Corrected by Antipurinergic Therapy.

(A) Control (Sal-Sal) synaptosome illustrating normal post-synaptic density (PSD) morphology (arrow), and normal electron lucency of the matrix (92,000× magnification; scale bar = 200 µm). (B) Treated controls (Sal-Sur) with an included mitochondrion (“m”; scale bar = 500 µm). (C) Untreated MIA (PIC-Sal) with an included mitochondrion (“m”) and malformed, hypomorphic PSD (arrow; scale bar = 500 µm). Note the abnormal accumulation of electron-dense matrix material. (D) Treated MIA (PIC-Sur) with restoration of near-normal PSD morphology (arrow), an included mitochondrion (“m”), and reduction in abnormal accumulations of electron-dense matrix material within the synaptosomes (scale bar = 500 µm). Representative fields from n = 3–4 males per group; age = 16 weeks.

Figure 5. Cerebral Mitochondrial Respiratory Chain Subunit Mass was Unchanged in the MIA Model.

Cerebral mitochondria were isolated by Percoll gradient centrifugation and quantified by Western Analysis. Each lane contains the mitochondria from 2–3 males isolated at 16-weeks of age (n = 4–5 per group).

Figure 6. Cerebral Mitochondrial Respiratory Chain Hyperfunction Abnormalities Were Corrected by Antipurinergic Therapy.

(A) Mitochondrial Respiratory Chain Complex I Enzymatic Activity was Increased 34% by gestational poly(IC) exposure and corrected by suramin treatment (Sal-Sal = 152+/−2.3 U/CS; PIC-Sal = 205+/−2.9 U/CS; one-way ANOVA F(3,12) = 137; p<0.0001; n = 4–5 males per group; age = 16 weeks). (B) Complex IV Activity was increased 53% by gestational poly(IC) exposure and corrected by suramin treatment (Sal-Sal = 0.319+/−0.045 U/CS; PIC-Sal = 0.492+/−0.018; one-way ANOVA F(3,12) = 8.9; p = 0.0022; n = 4–5 males per group; age = 16 weeks). Values are expressed as mean +/− SEM.

Figure 7. Cerebral Synaptosomal Purinergic Receptors were Downregulated in the MIA Model and Restored to Normal by Antipurinergic Therapy.

(A) Western Analysis of Metabotropic P2Y2 and Ionotropic P2X7 receptors. Each lane contains the synaptosomes from 2–3 males isolated at 16-weeks of age (n = 4–5 per group). (B) P2Y2 receptor expression was decreased by over 50% by gestational poly(IC) exposure and normalized by suramin treatment (Sal-Sal = 100+/−7.3%; Sal-Sur = 62+/−4.6%; PIC-Sal = 48+/−4.7%; PIC-Sur = 84+/−4.7%; one-way ANOVA F(3,12) = 18.1; p<0.0001; n = 4–5 males per group). (C) P2X7 receptor expression was decreased over 50% by gestational poly(IC) exposure and normalized by suramin treatment (Sal-Sal = 100+/−2.2%; Sal-Sur = 39+/−12%; PIC-Sal = 47+/−0.5%; PIC-Sur = 81+/−1.5%; one-way ANOVA F(3,12) = 23.2; p<0.0001; n = 4–5 males per group). Post-synaptic density 95 (PSD95) protein was used as a loading control. Values are expressed as mean +/− SEM.

Figure 8. Cerebral Synaptosomal Hypophosphorylation of Extracellular Response Kinase 1 and 2 (ERK1/2), Calcium-Calmodulin Kinase II (CAMKII), and Downregulation Fragile X Protein (FMRP) in the MIA Model Were Corrected, and Nicotinic Acetylcholine Receptor subunit α7 (nAchRα7) Expression was Increased by Antipurinergic Therapy.

(A) Western analysis of phosphorylated ERK1/2 (pERK1/2^{Thr202/Tyr204}), total ERK1/2, phosphorylated CAMKII (pCAMKII^Thr286), total CAMKII (CAMKIIpan), FMRP (Fragile X Syndrome protein), Nicotinic Acetylcholine Receptor subunit α7 (nAchRα7), and PSD95 (post-synaptic density protein 95) was used as a loading control for synaptosomes. Each lane contains the synaptosomes from 2–3 males isolated at 16-weeks of age (n = 4–5 per group).

Figure 9. Quantitative Analysis of the Extracellular Response Kinase 1 and 2 (ERK1/2), Calcium-Calmodulin Kinase II (CAMKII), Fragile X Protein (FMRP), and the Nicotinic Acetylcholine Receptor subunit α7 (nAchRα7) in Cerebral Synaptosomes in the MIA Model.

(A) ERK1 (MAPK3) and ERK2 (MAPK1) phosphorylation was decreased by over 90% by gestational poly(IC) exposure and normalized by suramin treatment (Sal-Sal = 100+/−5%; Sal-Sur = 6.2+/−0.8%; PIC-Sal = 9.9+/−2.6%; PIC-Sur = 84+/−2.3%; one-way ANOVA F(3,12) = 241; p<0.0001; n = 4–5 males per group; age = 16 weeks). (B) CAMKII phosphorylation was decreased by over 50% by gestational poly(IC) exposure and normalized by suramin treatment (Sal-Sal = 100+/−3%; Sal-Sur = 29+/−4.5%; PIC-Sal = 49+/−1.9%; PIC-Sur = 106+/−13%; one-way ANOVA F(3,12) = 25; p<0.0001; n = 4–5 males per group; age = 16 weeks). (C) Fragile X Mental Retardation Protein (FMRP) expression was decreased by over 40% by gestational poly(IC) exposure and normalized by suramin treatment (Sal-Sal = 100+/−8%; Sal-Sur = 56+/−1.2%; PIC-Sal = 55+/−2.5%; PIC-Sur = 89+/−6.5%; one-way ANOVA F(3,12) = 17.7; p<0.0001; n = 4–5 males per group; age = 16 weeks). (D) Nicotinic Acetylcholine Receptor subunit α7 (nAchRα7) expression was increased by over 75% by suramin treatment. (one-way ANOVA F(3,12) = 14.1; Sal-Sal = 100+/−9%; PIC-Sur = 176+/−18%; p<0.001; n = 4–5 males per group; age = 16 weeks; Newman-Keuls post test). Values are expressed as mean +/− SEM.

Figure 10. Purkinje Cell Dropout Was Prevented by Antipurinergic Therapy.

(A) Mosaic reconstruction of a representative parasagittal section of the cerebellar vermis in an untreated MIA animal (PIC-Sal). Sections were stained for calbindin (green) and neuN (red). Purkinje cells are the bright, large neurons located at the margins of the molecular (green) and granular (red) layers of the cerebellum. Lobules III though X are indicated. MIA animals at 16 weeks of age, showed patchy loss of Purkinje cells that was most striking in lobules VI and VII. (B) Higher magnification of lobule VII in a control (Sal-Sal) animal illustrating normal Purkinje cell numbers. (C) Higher magnification of lobule VII in MIA illustrating nearly complete Purkinje cell dropout, with scattered cells at the boundary between lobules VII and VIII in an animal exposed to poly(IC) during gestation. (D) Quantitation of Lobule VII Purkinje Cells. Animals exposed to poly(IC) during gestation had a 63% reduction in Purkinje cell numbers. This was prevented by suramin treatment (10–20 mg/kg ip qWeek) started at 6 weeks of age (Sal-Sal = 16.0+/−1.8 cells/mm; Sal-Sur = 19.6+/−2.5; PIC-Sal = 5.8+/−1.6, PIC-Sur = 20.4+/−5.4; one-way ANOVA F(3,13) = 5.3; p = 0.013; Newman-Keuls post hoc test; n = 4–5 males per group; age = 16 weeks). Values are expressed as mean +/− SEM.