Dysregulated balance of D- and L-amino acids modulating glutamatergic neurotransmission in severe spinal muscular atrophy

Abstract

Spinal muscular atrophy (SMA) is a neuromuscular disorder caused by reduced expression of the survival motor neuron (SMN) protein. In addition to motor neuron survival, SMN deficiency affects the integrity and function of afferent synapses that provide glutamatergic excitatory drive essential for motor neuron firing and muscle contraction. However, it is unknown whether deficits in the metabolism of excitatory amino acids and their precursors contribute to neuronal dysfunction in SMA. To address this issue, we measured the levels of the main neuroactive D- and L-amino acids acting on glutamatergic receptors in the central nervous system of SMN∆7 mice as well as the cerebrospinal fluid (CSF) of SMA patients of varying severity before and after treatment with the SMN-inducing drug Nusinersen. Our findings reveal that SMN deficiency is associated with disruption of glutamate and serine metabolism in the CSF of severe SMA patients, including decreased concentration of L-glutamate, which is partially corrected by Nusinersen therapy. Moreover, we identify dysregulated l-glutamine/L-glutamate ratio as a shared neurochemical signature of altered glutamatergic synapse metabolism that implicates neuron-astrocyte dysfunction in both severe SMA patients and mouse models. Lastly, consistent with hypo-glutamatergic neurotransmission in SMA, we show that daily supplementation with the NMDA receptor co-agonist d-serine improves neurological deficits in SMN∆7 mice. Altogether, these findings provide direct evidence for central dysregulation of D- and L-amino acid metabolism linked to glutamatergic neurotransmission in severe SMA and have potential implications for treating this neurological disorder.

Keywords: Central nervous system; Cerebrospinal fluid; Glutamatergic neurotransmission; NMDA receptors; Nusinersen; Spinal muscular atrophy; d-serine.

Fig. 1.

Levels of neuroactive amino acids in the CSF of SMA patients and control individuals. (A) Representative chromatogram showing the peaks of l-aspartate (L-Asp), L-glutamate (L-Glu), d-serine (D-Ser), l-serine (L-Ser), and l-glutamine (L-Gln) in CSF of SMA1 patients. (B-H) Levels of L-Glu (B), L-Gln (C), L-Gln/L-Glu ratio (D), L-Asp (E), L-Ser (F), D-Ser (G) and D-Ser/total Ser percentage ratio (H) in the indicated cohorts of SMA1, SMA2 and SMA3 patients as well as control individuals. Data are shown as violin plots representing the median with interquartile range (IQR). *P < 0.05, **P < 0.01(Mann-Whitney test with Bonferroni’s correction). Dots represent values from each individual analyzed.

Fig. 2.

Effect of Nusinersen on the levels of neuroactive amino acids in the CSF of SMA1 patients. (A) Schematic representation of the timeline of intrathecal Nusinersen administration and CSF collection in SMA patients. (B-H) Levels of L-glutamate (L-Glu) (B), L-glutamine (L-Gln) (C), L-glutamine/L-glutamate (L-Gln/L-Glu) ratio (D), L-aspartate (L-Asp) (E), L-serine (L-Ser) (F), D-serine (D-Ser) (G) and D-serine/total serine (D-Ser/total Ser) percentage ratio (H) in the CSF of SMA1 patients before treatment (T0, n = 18) and at the time of the sixth (T302, n = 18) injection of Nusinersen. *P < 0.05, **P < 0.01, compared to T0 (Wilcoxon matched-pairs signed ranks test). Data are shown as violin plots representing the median with interquartile range (IQR). Dots represent values from individual SMA1 patients.

Fig. 3.

Analysis of neuroactive amino acid levels in the brain and spinal cord of SMNΔ7 mice at early and late symptomatic stage of the disease. A) Schematic model of the tripartite glutamatergic synapse showing the main localization of the amino acids analyzed in this study. Image created with BioRender.com (www.biorender.com). Abbreviations: PHGDH: phosphoglycerate dehydrogenase; PSAT: phosphoserine aminotransferase; PSPH: phosphoserine phosphatase; DAAO: D-amino acid oxidase; SR: serine racemase; GLS: glutaminase; GS: glutamine synthetase; ASCT1: alanine, serine, cysteine transporter 1; ASC1: alanine, serine, cysteine transporter 1; GLT-1: glutamate transporter 1; SNAT: sodium-coupled neutral amino acid transporter; GLAST: glutamate aspartate transporter; NMDAR: N-methyl-d-aspartate receptor; AMPAR: alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor; VGLUT: vesicular glutamate transporter. B) Representative chromatogram showing the peaks of l-aspartate (L-Asp), L-glutamate (L-Glu), d-serine (D-Ser), l-serine (L-Ser), and l-glutamine (L-Gln) in the brain homogenate of SMNΔ7 mice. (C-F) Levels of L-Glu, L-Gln and L-Gln/L-Glu ratio in the brain and spinal cord of wild type (WT) and SMNΔ7 mice at postnatal day 3 (P3), and P11. The average amounts of amino acids detected were normalized for mg of total proteins. Dots represent values from individual mice. Amino acid levels are expressed as violin plots representing median with interquartile range (IQR) and analyzed by Mann-Whitney test (*P < 0.05, **P < 0.01, compared to age-matched WT mice). (G,H) Amino acid levels were also analyzed as two-way ANOVA, followed by Tukey’s multiple comparisons test (**P < 0.01, **P < 0.0001, compared to age-matched WT mice; ^##P < 0.01, ^###P < 0.0001, compared to genotype-matched P3 mice). Amino acid levels were shown as scatter dot plots representing median with interquartile range (IQR) while dots represent values from individual mice.

Fig. 4.

SMN-independent amelioration of motor function by d-serine supplementation in SMA mice. (A) Representative chromatogram showing the peaks of d-serine (D-Ser) and l-serine (L-Ser) in the spinal cord homogenate of SMNΔ7 mice after a single injection of vehicle (black line) or 500 mg/kg D-Ser treatment (blu line) at P3. (B-C) D-Ser levels and D-Ser/toral Ser ratio in the spinal cord of WT mice treated with a single injection of 500 mg/kg d-serine compared to vehicle-treated mice at P3. *P < 0.05, compared to vehicle (Mann-Whitney test) (D) Western blot analysis of SMN protein levels in spinal cords from WT and either untreated or D-Ser-treated SMNΔ7 mice at P11. GAPDH was probed as a loading control. Three independent mice per group were analyzed. (E) Quantification of SMN levels normalized to GAPDH from the data in (D). Ordinary one-way ANOVA with Šídák’s multiple comparisons test. ****P < 0.0001, ns: not significant. Values are means and SEM. (F-H) Analysis of righting time (F), body weight (G), and survival (H) in WT mice (n = 19) and either untreated (n = 17) or D-Ser-treated (n = 16) SMNΔ7 mice. Two-way ANOVA with Tukey’s multiple comparisons test. ****P < 0.0001, ***P < 0.001, *P < 0.05. Values are means and SEM (see also Suppl. Table 6 for raw data).