Genetic Downregulation of the Metabotropic Glutamate Receptor Type 5 Dampens the Reactive and Neurotoxic Phenotype of Adult ALS Astrocytes

Abstract

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by progressive degeneration of motor neurons (MNs). Astrocytes display a toxic phenotype in ALS, which results in MN damage. Glutamate (Glu)-mediated excitotoxicity and group I metabotropic glutamate receptors (mGluRs) play a pathological role in the disease progression. We previously demonstrated that in vivo genetic ablation or pharmacological modulation of mGluR5 reduced astrocyte activation and MN death, prolonged survival and ameliorated the clinical progression in the SOD1^G93A mouse model of ALS. This study aimed to investigate in vitro the effects of mGluR5 downregulation on the reactive spinal cord astrocytes cultured from adult late symptomatic SOD1^G93A mice. We observed that mGluR5 downregulation in SOD1^G93A astrocytes diminished the cytosolic Ca²⁺ overload under resting conditions and after mGluR5 simulation and reduced the expression of the reactive glial markers GFAP, S100β and vimentin. In vitro exposure to an anti-mGluR5 antisense oligonucleotide or to the negative allosteric modulator CTEP also ameliorated the altered reactive astrocyte phenotype. Downregulating mGluR5 in SOD1^G93A mice reduced the synthesis and release of the pro-inflammatory cytokines IL-1β, IL-6 and TNF-α and ameliorated the cellular bioenergetic profile by improving the diminished oxygen consumption and ATP synthesis and by lowering the excessive lactate dehydrogenase activity. Most relevantly, mGluR5 downregulation hampered the neurotoxicity of SOD1^G93A astrocytes co-cultured with spinal cord MNs. We conclude that selective reduction in mGluR5 expression in SOD1^G93A astrocytes positively modulates the astrocyte reactive phenotype and neurotoxicity towards MNs, further supporting mGluR5 as a promising therapeutic target in ALS.

Keywords: Grm5 genetic ablation; SOD1G93A mice; adult mouse spinal cord astrocytes; amyotrophic lateral sclerosis; mGlu5 receptor.

Figure 1

mGluR5 expression in spinal cord astrocytes cultured from adult Grm5^−/+, WT, SOD1^G93A, and SOD1^G93AGrm5^−/+ mice. (A) Representative Western blot (WB) showing immunoreactive mGluR5 bands. (B) Quantification of WB mGluR5 densitometric signals. Protein band density was normalized to GAPDH, as a housekeeping protein. Data are means ± s.e.m of n = 3 independent experiments. * p < 0.05 and *** p < 0.001 vs. WT astrocytes; ### p < 0.001 vs. SOD1^G93A astrocytes (F_{(3, 8)} = 43.43; one-way ANOVA followed by Tukey’s multi-comparison test). (C) Representative confocal microscopy immunocytochemical images of mGluR5 (green fluorescence) and lectin (red fluorescence). Scale bar 100 µm. Grm5^−/+, WT, SOD1^G93A, and SOD1^G93AGrm5^−/+ spinal cord astrocytes were fixed with paraformaldehyde and incubated with the primary antibodies for mGluR5 and lectin, without membrane permeabilization, and subsequently with fluorescent secondary antibodies. Images were acquired by confocal microscopy. (D) Quantitative representation of mGluR5 expression was calculated as the relative fluorescence intensity of mGluR5 co-localized with lectin, that labels the plasma membrane. Data are means ± s.e.m of n = 8 independent experiments. **** p < 0.0001 vs. WT astrocytes; #### p < 0.0001 vs. SOD1^G93A astrocytes (F_(3,28) = 134.3; one-way ANOVA followed by Tukey’s multi-comparison test).

Figure 2

Ca²⁺ concentration [Ca²⁺]i under basal and stimulated conditions in spinal cord astrocytes cultured from adult WT, Grm5^−/+, SOD1^G93A, and SOD1^G93AGrm5^−/+ mice. Astrocytes were labeled with the fluorescent dye Fura2-AM. [Ca²⁺]_i was measured by ratiometric spectrofluorophotometry in basal conditions and after stimulation with the selective group I metabotropic glutamate receptor agonist 3,5-DHPG (30 μM). Results are means ± s.e.m of n = 6–9 independent experiments and are expressed as nanomolar (nM) Ca²⁺ concentration [Ca²⁺]i. §§§§ p < 0.0001 stimulated vs. the respective basal conditions; **** p < 0.0001 vs. WT astrocytes, under basal or stimulated conditions; #### p < 0.0001 vs. SOD1^G93A astrocytes under basal or stimulated conditions (F_(1,58) = 207.2 and F_(3,58) = 240.5; two-way ANOVA followed by Tukey’s multi-comparison test).

Figure 3

Expression and cellular localization of the astrocyte activation markers GFAP, vimentin, and S100β in spinal cord astrocytes cultured from adult WT, Grm5^−/+, SOD1^G93A, and SOD1^G93AGrm5^−/+ mice. (A) Representative Western blots (WBs) for GFAP, vimentin and S100β (B–D). Densitometric quantification of WB signals of (B) GFAP, (C) vimentin and (D) S100β. Protein band density was normalized to GAPDH as a housekeeping protein. Data are means ± s.e.m of n = 3 independent experiments. ** p < 0.01, *** p < 0.001 and **** p < 0.0001 vs. WT astrocytes; # p < 0.05, ## p < 0.01 and ### p < 0.001 vs. SOD1^G93A astrocytes, (F_(3,8) = 77.99, F_(3,8) = 131.5 and F_(3,8) = 80.76 for GFAP, vimentin and S100β, respectively; one-way ANOVA followed by Tukey’s multi-comparison test). (E–G) Representative confocal microscopy immunocytochemical images of GFAP (E), vimentin (F) and S100β (G) (green fluorescence) and GAPDH (red fluorescence). Grm5^−/+, WT, SOD1^G93A and SOD1^G93AGrm5^−/+ spinal cord astrocytes were fixed, permeabilized and incubated with appropriate primary and fluorescent secondary antibodies. Images were acquired by confocal microscopy. Scale bar: 100 µm. Quantitative representation of GFAP (H), vimentin (I), and S100β (J) expression, calculated as the relative fluorescence intensity of the protein of interest co-localized with the reference protein GAPDH. Data are means ± s.e.m of n = 7–8 independent experiments. ** p < 0.01 and **** p < 0.0001 vs. WT astrocytes; #### p < 0.0001 vs. SOD1^G93A astrocytes (F_(3,28) = 190.5, F_(3,28) = 44.80 and F_(3,28)= 49.33 for GFAP, vimentin and S100β, respectively; one-way ANOVA followed by Tukey’s multi-comparison test).

Figure 4

Effect of in vitro exposure to an mGluR5 antisense oligonucleotide and the mGluR5 selective negative allosteric modulator CTEP in spinal cord astrocytes cultured from adult SOD1^G93A mice. (A) RT-qPCR quantitative analyses for Grm5 expression in untreated, scramble- or ASO-treated (20 µM) SOD1^G93A spinal cord astrocytes. Data are means ± s.e.m of n = 3 independent experiments run in triplicate. # p < 0.05 vs. untreated SOD1^G93A astrocytes; $ p < 0.05 vs. scramble-treated SOD1^G93A astrocytes (F_(2,6) = 11.15; one-way ANOVA followed by Tukey’s multi-comparison test). (B,C) Representative confocal microscopy immunocytochemical images of (B) GFAP (green fluorescence) and GAPDH (red fluorescence) and (C) S100β (green fluorescence) and GAPDH (red fluorescence) in untreated, scramble- and ASO-treated SOD1^G93A astrocytes. Scale bar: 100 µm. Astrocytes were labeled with appropriate primary and fluorescent secondary antibodies, and the images were acquired by confocal microscopy. Quantitative representation of GFAP (D) and S100β (E) expression, calculated as the relative fluorescence intensity of the protein of interest co-localized with the reference protein GAPDH. Data are means ± s.e.m of n = 8 independent experiments. #### p < 0.0001 vs. untreated SOD1^G93A astrocytes and $$ p < 0.0001 vs. scramble-treated SOD1^G93A astrocytes (F_(2,21) = 24.39 and F_(2,21) = 60.46 for GFAP and S100β, respectively; one-way ANOVA followed by Tukey’s multi-comparison test). (F,G) Representative confocal microscopy immunocytochemical images of (F) GFAP (green fluorescence) and GAPDH (red fluorescence) and (G) S100β (green fluorescence) and GAPDH (red fluorescence) in SOD1^G93A astrocytes exposed for 7 days to DMSO or CTEP (100 nM). Scale bar: 100 µm. (H,I) Quantitative representation of GFAP (H) and S100β (I) protein expression, calculated as described above. Data are means ± s.e.m of n = 6 independent experiments. ^#### p < 0.0001 vs. SOD1^G93A astrocytes treated with DMSO (F_(2,15) = 209.8 and F_(2,15) = 234.2 for GFAP and S100β, respectively; one-way ANOVA followed by Tukey’s multi-comparison test).

Figure 5

NLRP-3 inflammasome complex expression, cytokines and glutamate release in the culture medium of spinal cord astrocytes cultured from adult WT, SOD1^G93A and SOD1^G93AGrm5^−/+ mice. (A–C) Representative Western blot (WB) immunoreactive NLRP-3 (A) and IL-1β, IL-6, and TNF-α (C) bands. (B,D–F) Quantitative representation of WB densitometric signals of (B) NLRP-3, (D) IL-1β, (E) IL-6, and (F) TNF-α. Protein band density was normalized to GAPDH as a housekeeping protein. Data are means ± s.e.m of n = 3 independent experiments. ** p < 0.01, *** p < 0.001 and **** p < 0.0001 vs. WT astrocytes; # p < 0.05, ## p < 0.01 and ### p < 0.001 vs. SOD1^G93A astrocytes (F_(3,8) = 55.83, F_(2,6) = 23.18, F_(2,6) = 106.2 and F_(2,6) = 38.49 for NLRP-3, IL-1β, IL-6 and TNF-α respectively; one-way ANOVA followed by Tukey’s multi-comparison test). (G–I) Enzyme-linked immunosorbent (ELISA) assay of the (G) IL-1β, (H) IL-6, and (I) TNF-α released in the astrocyte culture medium. Cell culture medium was collected after 24 h, and the inflammatory cytokine content was measured with specific ELISA kits and expressed as pmol per mg of astrocyte’s total protein content in each well. Data means ± s.e.m of n = 6 independent experiments. ** p < 0.01 and **** p < 0.0001 vs. WT astrocytes; #### p < 0.0001 vs. SOD1^G93A astrocytes (F_(2,15) = 1563, F_(2,15) = 2426 and F_(2,15)= 251.4, for IL-1β, IL-6 and TNF-α, respectively; one-way ANOVA followed by Tukey’s multi-comparison test). (J) Glu release in the astrocyte culture medium. Cell culture medium was replaced with Hepes-buffered physiological medium, which was collected after 4 h. Glu content was measured by HPLC after orthophthaldialdehyde derivatization and fluorometric detection. The physiological medium without cells contained 0.01 ± 0.002 µM of Glu (not shown). Glu is expressed as micromolar (µM) concentrations in each well, containing an average of 1 × 10⁵ cells. Data are means ± s.e.m of n = 16–27 wells (independent biological replicates). **** p < 0.0001 vs. WT astrocytes (F_(2,57) = 40.70; one-way ANOVA followed by Tukey’s multi-comparison test).

Figure 6

Energy metabolism in spinal cord astrocytes cultured from adult WT, SOD1^G93A and SOD1^G93AGrm5^−/+ mice. (A,B) Oxygen consumption. After permeabilization with 0.03% digitonin, WT, SOD1^G93A, and SOD1^G93AGrm5^−/+ astrocytes were resuspended in a respiration medium and stimulated (A) with pyruvate (10 mM) + malate (5 mM) and ADP (0.1 mM) to evaluate the cellular respiration through the Complex I, III and IV pathway or (B) with succinate (20 mM) and ADP (0.1 mM) to investigate the activity of the Complex II, III, and IV pathway. The respiratory rate was expressed as nmol consumed oxygen/min/10⁶ cells. Data are means ± s.e.m of n = 5 independent experiments run in triplicate. ** p < 0.01 and **** p < 0.0001 vs. WT astrocytes; #### p < 0.0001 vs. SOD1^G93A astrocytes (pyruvate + malate: F_(2,12) = 59.22; succinate: F_(2,12) = 173.5; one-way ANOVA followed by Tukey’s multi-comparison test). (C,D) ATP synthesis by F_o-F₁ ATP synthase evaluated after stimulation (C) with pyruvate (10 mM) + malate (5 mM) or (D) with succinate (20 mM), and ADP (0.1 mM) in both conditions. Data are expressed as nmol ATP produced/min/10⁶ cells and are means ± s.e.m of n = 5 independent experiments. *** p < 0.001 and **** p < 0.0001 vs. WT astrocytes; #### p < 0.0001 vs. SOD1^G93A astrocytes (pyruvate and malate: F_(2,12) = 74.56; succinate: F_(2,12) = 344.6; one-way ANOVA followed by Tukey’s multi-comparison test). (E,F) P/O ratio as an OxPhos efficiency marker. The P/O was calculated as the ratio between the concentration of the produced ATP and the amount of consumed oxygen in the presence of (E) pyruvate (10 mM) + malate (5 mM) and ADP (0.1 mM) or (F) succinate (20 mM) and ADP (0.1 mM). Data are means ± s.e.m of n = 5 independent experiments run in triplicate. * p < 0.05, ** p < 0.01 and **** p < 0.0001 vs. WT astrocytes; #### p < 0.0001 vs. SOD1^G93A astrocytes (F_(2,12) = 94.92 for pyruvate and malate; succinate: F_(2,12) = 88.25; one-way ANOVA followed by Tukey’s multi-comparison test). (G) Lactate dehydrogenase (LDH) activity. LDH activity is expressed as international milliunits (mU/mg), corresponding to the nanomoles of substrate catalyzed in 1 min per mg of protein. Data are means ± s.e.m of n = 5 independent experiments, run in triplicate. **** p < 0.0001 vs. WT astrocytes; #### p < 0.0001 vs. SOD1^G93A astrocytes (F_(2,14) = 129; one-way ANOVA followed by Tukey’s multi-comparison test). (H) Lactate and (I) glucose concentrations were assayed spectrophotometrically in the culture medium. Data are expressed as mM concentration per 10⁶ cells and are means ± s.e.m of n = 5 independent experiments. **** p < 0.0001 vs. WT astrocytes; #### p < 0.0001 vs. SOD1^G93A astrocytes (F_(2,12) = 47.43; one-way ANOVA followed by Tukey’s multi-comparison test).

Figure 7

Survival of WT and SOD1^G93A MNs co-cultured with adult astrocytes from WT, SOD1^G93A and SOD1^G93AGrm5^−/+ mice. (A–D) Representative phase-contrast microscopy images (100×) of (A) WT MNs co-cultured with WT astrocytes, (B) WT MNs plated with SOD1^G93A astrocytes, (C) SOD1^G93A MNs plated with SOD1^G93A astrocytes and (D) SOD1^G93A MNs plated with SOD1^G93AGrm5^−/+ astrocytes; scale bar: 50 µm. (E) Quantification of MN viability. MNs were isolated from the spinal cord of WT and SOD1^G93A E13.5 mouse embryos and seeded on mature adult astrocyte cultures as described above. MNs were counted in a 1 cm² area starting from day 4 after seeding, three times a week, for 14 days and expressed as percent (%) MN survival vs. the respective number of MNs counted on day 4. Data are means ± SEM of n = 6–9 independent experiments. * p < 0.05, *** p < 0.001 and **** p < 0.0001 vs. WT MNs co-cultured with WT astrocytes; ## p < 0.01 and ### p < 0.001 vs. SOD1^G93A MNs co-cultured with SOD1^G93A astrocytes; §§ p < 0.01 vs. WT MNs co-cultured with SOD1^G93A astrocytes (day 6: F_(3,29) = 14.41; day 8: F_(3,29) = 8.820; day 10: F_(3,29) = 5.169; day 12: F_(3,29) = 6.032; day 14: F_(3,29) = 6.604; one-way ANOVA followed by Tukey’s multi-comparison test).