Mutation of ataxia-telangiectasia mutated is associated with dysfunctional glutathione homeostasis in cerebellar astroglia

Abstract

Astroglial dysfunction plays an important role in neurodegenerative diseases otherwise attributed to neuronal loss of function. Here we focus on the role of astroglia in ataxia-telangiectasia (A-T), a disease caused by mutations in the ataxia-telangiectasia mutated (ATM) gene. A hallmark of A-T pathology is progressive loss of cerebellar neurons, but the mechanisms that impact neuronal survival are unclear. We now provide a possible mechanism by which A-T astroglia affect the survival of cerebellar neurons. As astroglial functions are difficult to study in an in vivo setting, particularly in the cerebellum where these cells are intertwined with the far more numerous neurons, we conducted in vitro coculture experiments that allow for the generation and pharmacological manipulation of purified cell populations. Our analyses revealed that cerebellar astroglia isolated from Atm mutant mice show decreased expression of the cystine/glutamate exchanger subunit xCT, glutathione (GSH) reductase, and glutathione-S-transferase. We also found decreased levels of intercellular and secreted GSH in A-T astroglia. Metabolic labeling of l-cystine, the major precursor for GSH, revealed that a key component of the defect in A-T astroglia is an impaired ability to import this rate-limiting precursor for the production of GSH. This impairment resulted in suboptimal extracellular GSH supply, which in turn impaired survival of cerebellar neurons. We show that by circumventing the xCT-dependent import of L-cystine through addition of N-acetyl-L-cysteine (NAC) as an alternative cysteine source, we were able to restore GSH levels in A-T mutant astroglia providing a possible future avenue for targeted therapeutic intervention.

Keywords: astroglia; ataxia-telangiectasia; ataxia-telangiectasia mutated; cerebellum; glutathione; neuronal survival; xCT.

FIGURE 1

Effect of A–T and WT astroglial on cerebellar neuron survival and neurite extension. (A, B) Quantification of the total number of A–T PNs (A) or A–T GCs (B) cocultured on astroglia monolayers derived from either WT (white bar) of A–T (dark bar) cerebellum and grown for 3 days. Cells were labeled with DAPI to determine the number of live cells/well as defined by the absence of pyknotic nuclei. Shown are the average numbers of cells from independent experiments. A total of approximately 200 cells/well for PNs and 2000 cells/well for CNS were plated on day 0. Error bars represent SEM. Quantification was conducted using Student’s t-test. *P = 0.014 with n = 10 for PNs and P = 0.02 with n = 6 for CGNs. (C, D) The left panels show representative images of A–T CGNs (labeled with anti-β-III tubulin in red), grown on WT (C) or A–T (D) astrocyte monolayers (astrocytes are labeled with anti-GFAP in green shown on the right panel). All cultures were also labeled with DAPI to visualize nuclei. (Note: the right panel represents the triple-labeled composite. To clearly visualize neurons grown on astrocytes, the green astrocyte labeling is omitted in the left panel). (E, F) Quantification of the average length of neurites extended from (E) A–T and (F) WT CGNs cultured on WT or A–T astroglia. An average of 500 neurons from three different dissections were analyzed. Error bars represent standard error of the mean (SEM) ***P < 0.005 with n = 4 by Student’s t-test.

FIGURE 2

Effect of A–T astroglia conditioned medium on neuronal survival and outgrowth. (A, B) For these experiments, we cultured A–T (A) and WT (B) CGNs on poly-L-lysine (PLL)-coated wells in the presence of ACM and determined the number of surviving cells after 3 days in culture. ACM collected from WT astroglia served as positive control. Minimal media devoid of growth factors serves as negative control. **P < 0.01 by One-way ANOVA followed by Bonferroni posttest. n = 3 for WT CGNs and n = 4 for A–T CGNs. (C) Images show CGNs grown in culture and labeled with anti-βIII tubulin in WT ACM and A–T ACM and represent color inversions to enhance contrast. (D, E) Quantification of average neurite length of A–T CGNs (D) and WT CGNs (E) cultured in ACM from A–T or WT astroglia. *P < 0.05 with n = 3 by one-way ANOVA followed by Bonferroni post-test.

FIGURE 3

Decreased GSH content of A–T astroglia reduces neuronal survival. (A, B) Cerebellar A–T astroglia show a 40% decrease in (A) cellular and 85% decrease in (B) extracellular GSH relative to WT controls as measured by LC-MS/MS. **P < 0.01, ***P < 0.001 with n = 6 for cell lysates and n = 9 for ACM by Student’s t-test. In vitro survival of (C) A–T and (D) WT CGNs was increased by the addition of exogenous 250 μM GSH to A–T ACM. *P < 0.05, **P < 0.01, ***P < 0.001 with n = 3 by two-way ANOVA followed by Bonferroni post-test.

FIGURE 4

GSH pathway proteins and mRNA expression levels are altered in A–T cerebellar astroglia and human A–T cerebellar tissue. (A) Representative immunoblots and (B) expression levels of key GSH pathway proteins in A–T cerebellar astroglia relative to WT controls. A–T cerebellar astroglia show reductions in xCT, GST-μ, and GR. (C) qPCR analysis shows decreased GR and markedly decreased xCT mRNA expression in A–T cerebellar astroglia. (D) Representative immunoblots of human cerebellar tissue show a reduction in GPx1, GR, GST-μ, and xCT in the A–T cerebellar sample relative to an age matched control. *P < 0.05, **P < 0.01, ***P < 0.001 by Mann–Whitney test. For immunoblot analysis, n = 3 (xCT, GSS, MRP1), n = 6 (GPx1), n = 8 (GST-μ), n = 9 (GR), and n = 10 (γGCS). For RNA expression analysis, n = 4.

FIGURE 5

A–T cerebellar astroglia exhibit decreased L-cystine incorporation into GSH. (A) Labeled ¹³C-L-cystine is transported into astroglia via xCT-mediated uptake and incorporated into GSH. The resulting stable isotope-labeled GSH (13C GSH) can be detected within the cell or the ACM using LC-MS/MS. (B) A–T cerebellar astroglia showed a significant decrease in 13C GSH in cell lysates after 18 h compared to WT controls. **P < 0.01 with n = 8 by Student’s t-test. (C) LC-MS/MS measurements of 13C GSH in ACM from A–T astroglia showed significantly less 13C GSH compared to control following an 18-h treatment with ¹³C6¹⁵N2 L-cystine. ***P < 0.001 with n = 8 by Student’s t-test.

FIGURE 6

NAC increases GSH production in A–T cerebellar astroglia. (A, B) Addition of the alternative cysteine source, NAC, to A–T cerebellar astroglia leads to a statistically significant increase in intracellular 12C GSH (A) and trend toward increased 13C GSH (B) content as measured by LC-MS/MS *P < 0.05, with n = 8 for control samples, n = 4 for NAC- and Trolox-treated astroglia by one-way ANOVA followed by Bonferroni posttest. (C, D) Addition of NAC or Trolox to mutant A–T astroglia led to a slight increase in in total GSH in the ACM at this time point but did not reach statistical significance. Addition of Trolox has no effect on the total GSH levels in the ACM (P > 0.05) with n = 8 for control samples, n = 4 for NAC- and Trolox-treated astroglia by one-way ANOVA followed by Bonferroni post-test.