Expanded-polyglutamine huntingtin protein suppresses the secretion and production of a chemokine (CCL5/RANTES) by astrocytes

Abstract

Huntington's disease (HD) is a hereditary neurological disease caused by expended CAG repeats in the HD gene, which codes for a protein called Huntingtin (Htt). The resultant mutant Huntingtin (mHtt) forms aggregates in neurons and causes neuronal dysfunction. In astrocytes, the largest population of brain cells, mHtt also exists. We report herein that astrocyte-conditioned medium (ACM) collected from astrocytes of R6/2 mice (a mouse model of HD) caused primary cortical neurons to grow less-mature neurites, migrate more slowly, and exhibit lower calcium influx after depolarization than those maintained in wild-type (WT) ACM. Using a cytokine antibody array and ELISA assays, we demonstrated that the amount of a chemokine [chemokine (C-C motif) ligand 5 (CCL5)/regulated on activation normal T cell expressed and secreted (RANTES)] released by R6/2 astrocytes was much less than that by WT astrocytes. When cortical neurons were treated with the indicated ACM, supplementation with recombinant CCL5/RANTES ameliorated the neuronal deficiency caused by HD-ACM, whereas removing CCL5/RANTES from WT-ACM using an anti-CCL5/RANTES antibody mimicked the effects evoked by HD-ACM. Quantitative PCR and promoter analyses demonstrated that mHtt hindered the activation of the CCL5/RANTES promoter by reducing the availability of nuclear factor kappaB-p65 and, hence, reduced the transcript level of CCL5/RANTES. Moreover, ELISA assays and immunocytochemical staining revealed that mHtt retained the residual CCL5/RANTES inside R6/2 astrocytes. In line with the above findings, elevated cytosolic CCL5/RANTES levels were also observed in the brains of two mouse models of HD [R6/2 and Hdh((CAG)150)] and human HD patients. These findings suggest that mHtt hinders one major trophic function of astrocytes which might contribute to the neuronal dysfunction of HD.

Figure 1.

ACM collected from R6/2 mice exhibited less activity in supporting neurite development of primary cortical neurons. A, B, Immunostaining of mHtt aggregates (green), GFAP (an astrocyte marker; red), and nucleus (DAPI; blue) in primary astrocytes (30 DIV) purified from the indicated mice. Representative merged images are shown. The corresponding images of mHtt, and GFAP and DAPI alone, are displayed in supplemental Figure S2A (available at www.jneurosci.org as supplemental material). The boxed region in B is enlarged to show the aggregates of mHtt. The arrow indicates mHtt aggregates in the nucleus. No aggregate was found in WT astrocytes. Scale bar, 20 μm. C–E, Immunocytochemical images of primary cortical neurons cultured with NB or the indicated ACM (50%, v/v) from 4 DIV. Cells were fixed at 8 DIV and stained with TUJ-1 (a neuron-specific marker; green) and DAPI (blue). F–H, The neurite length (F), the number of branchings (G), and the number of sproutings (H) of each neuron under the indicated condition were quantified using MetaMorph software. ^ap ≤ 0.001, compared with the NB group; ^bp ≤ 0.001, compared with the WT-ACM by Student's t test. Scale bar, 100 μm.

Figure 2.

Level of CCL5/RANTES was selectively reduced in ACM of R6/2 astrocytes. A, The cytokine antibody array revealed differential expression of CCL5/RANTES in ACM collected from WT or R6/2 cortical astrocytes as indicated. The intensities of the spots reflect the relative levels of the indicated proteins. Thick arrows, thin arrows, and the arrowhead mark the position of CCL5/RANTES, CCL2/MCP-1, and the positive control (IgG) on the same membrane, respectively. Quantization of CCL5/RANTES and CCL2/MCP-1 levels from cytokine array membranes is shown in the right panel. Data were normalized with an internal positive control on the same membrane and are presented as the mean ± SEM of four determinations from two independent batches of ACM and two sets of antibody arrays. *p < 0.05, compared with the WT-ACM, by Student's t test. B, Levels of CCL5/RANTES in the WT- and R6/2-ACM were determined by ELISA. To verify the specificity of the CCL5/RANTES signal, an anti-CCL5/RANTES neutralizing goat IgG (0.2 μg/ml) or a normal goat IgG (0.2 μg/ml) was added to the ACM as indicated. To reconstitute R6/2-ACM for further functional assays, recombinant mouse CCL5/RANTES (450 pg/ml) was added to R6/2-ACM and analyzed for the levels of CCL5/RANTES as shown. Values are presented as the mean ± SEM of 12 determinations from four different batches of ACM. Statistical analyses were conducted using Student's t test followed by the Mann–Whitney rank sum test. **p < 0.001, compared with WT-ACM containing the control IgG. C, Astrocytes from different brain areas of the indicated animals were prepared. Levels of CCL5/RANTES in the WT- and R6/2-ACM were determined by ELISA. Values are presented as the mean ± SEM from three different batches of ACM. Statistical analyses were conducted using Student's t test (*p < 0.05, compared with WT-ACM containing the control IgG).

Figure 3.

CCL5/RANTES in ACM was critical for neurite development of primary cortical neurons. A–D, Immunocytochemical images of primary cortical neurons cultured with the indicated ACM (50%, v/v) containing the control, IgG, the anti-CCL5/RANTES IgG, or recombinant CCL5/RANTES (450 pg/ml) from 4 DIV. Note that the addition of an anti-CCL5/RANTES IgG completely removed CCL5/RANTES from WT-ACM as shown in Figure 2. Cells were fixed at 8 DIV and stained with TUJ-1. E, F, The neurite length (E) and the number of branchings (F) of each neuron under the indicated condition were quantified using MetaMorph software. Totals of ∼63–131 cells were scored in the examined conditions. Data points represent the mean ± SEM from four independent experiments. ^ap ≤ 0.001, compared with the WT-ACM containing control goat IgG; ^bp ≤ 0.001, compared with the R6/2-ACM containing control IgG, by Student's t test. Scale bar, 100 μm.

Figure 4.

CCL5/RANTES in ACM mediated neuronal migration. Transwell migration assays for neuron migration in response to the indicated ACM containing the control IgG (Con-IgG), anti-CCL5/RANTES IgG, or recombinant CCL5/RANTES (450 pg/ml) were performed. After 18 h, the undersides of the transwell membranes were fixed and stained with crystal violet. A–D, Representative photographs. Migrating neurons (indicated by red arrows) appear purple and can be observed near the membrane pores (indicated by blue arrows). E, Numbers of migrating cells from five different fields of each condition were quantified. Data points represent the mean ± SEM of four independent experiments. Statistical analyses were conducted using Student's t test followed by the Mann–Whitney rank sum test. ^ap ≤ 0.01, compared with WT-ACM containing the control IgG; ^bp ≤ 0.05, compared with R6/2-ACM containing the control IgG.

Figure 5.

CCL5/RANTES in ACM maintained proper neuronal activity. Primary neurons (at 4 DIV) cultured in the indicated ACM containing the control IgG (0.2 μg/ml), the anti-CCL5/RANTES IgG (0.2 μg/ml), or the recombinant CCL5/RANTES (450 ng/ml) as indicated for 4 d. Cells were loaded with fura-2 (5 μm) and stimulated with 60 mm K⁺ (marked by the arrow). A, Representative traces depict the [Ca²⁺]_i elevation evoked by the high external K⁺ as indicated. B, The maximal calcium response after depolarization was quantified using Metafluor software. At least 100 cells for each condition were analyzed. Values are expressed as percentages of the calcium in WT astrocytes and represent the mean ± SEM of four independent experiments. ^ap < 0.05, compared with WT-ACM containing the control IgG; ^bp < 0.05, compared with R6/2-ACM containing the control IgG, by Student's t test followed by the Mann–Whitney rank sum test.

Figure 6.

mHtt reduced the gene expression of CCL5/RANTES in astrocytes via reducing the nuclear level of NFκB p65. A, The pcDNA3.1-(Q)₂₅-Htt-hrGFP or pcDNA3.1-(Q)₁₀₉-Htt-hrGFP construct was cotransfected with the indicated pGL3-CCL5 promoter construct at a ratio of 7:2 into primary WT astrocytes (30 DIV) for 48 h. The luciferase activity was measured and normalized to the protein content. Values are expressed as percentages of the promoter activity of pGL3-CCL5-P_{(−1064/−16)} in the presence of pcDNA3.1-(Q)₂₅-Htt-hrGFP and represent the mean ± SEM of four independent experiments. **p < 0.001, compared with that of the 25Q transfectant of each group, by paired t test. B, The pcDNA3.1-(Q)₂₅-Htt-hrGFP or pcDNA3.1-(Q)₁₀₉-Htt-hrGFP construct was cotransfected with pGL3-CCL5-P_{(−1064/−16)} and an expression construct of the indicated transcription factor (NFκB-p65, Sp1, p53, or HMG-1) at a ratio of 2:1:2 into primary WT astrocytes for 48 h. The luciferase activity was measured and normalized to the protein content. Values are expressed as percentages of the promoter activity of pGL3-CCL5-P_{(−1064/−16)} in the presence of pcDNA3.1-(Q)₂₅-Htt-hrGFP, but no overexpression of any transcription factor, and represent the mean ± SEM of four independent experiments. **p < 0.001, compared with that in the 25Q transfectant of each group; paired t test. C, The indicated Htt construct with 109 copies of polyQ was cotransfected with the WT or the indicated NFκB-null CCL5/RANTES promoter construct [pGL3-CCL5-P_{(−1064/−16)}-M1 or pGL3-CCL5-P_{(−1064/−16)}-M2] and an expression construct of NFκB-p65 as described above. The luciferase activity was measured and normalized to the protein content. Values are expressed as percentages of the promoter activity of pGL3-CCL5-P_{(−1064/−16)} in the presence of 25Q without any transcription factor and are presented as the mean ± SEM of three independent experiments. *p < 0.01, Student's t test. D, Nuclear fractions (22 μg per lane) collected from the indicated astrocytes were subjected to Western blot analysis. The level of nuclear NFκB-p65 in R6/2 astrocyte was normalized with an internal control (lamin A) and compared with that of WT astrocytes. Data are presented as the mean ± SEM from four independent experiments. **p < 0.001 compared with that in WT astrocytes, t test.

Figure 7.

mHtt suppressed the release of CCL5/RANTES from primary astrocytes. A, Immunostaining of CCL5/RANTES (green) and GFAP (red) was performed in primary astrocytes. Nuclei were stained with Hoechst 33258 (blue). CCL5/RANTES was enriched inside of R6/2 astrocytes (bottom) but not in WT astrocytes (top). Scale bar, 100 μm. B, WT astrocytes were transfected with DsRed:CCL5 along with the Htt-(Q)₂₅-hrGFP or Htt-(Q)₁₀₉-hrGFP construct at a ratio of 1:1 for 48 h. Cells were fixed, and the fluorescent signals were detected using confocal microscopy. Scale bars, 20 μm. C, D, Both WT and R6/2 astrocytes were treated with either 0.1% DMSO as a control or BFA (10 μg/ml) for 4 h to block secretion of any newly synthesized CCL5/RANTES protein. Intracellular CCL5/RANTES levels were assessed by immunocytochemical staining using an anti-CCL5/RANTES. CCL5/RANTES accumulation was enhanced by BFA in WT astrocytes, but not in R6/2 astrocytes, as shown in supplemental Figure S4 (available at www.jneurosci.org as supplemental material). C, CCL5/RANTES released extracellularly from astrocytes was collected and detected using a CCL5/RANTES ELISA assay. D, The relative CCL5/RANTES fluorescence intensities of both vehicle- and BFA-treated WT and R6/2 astrocytes were quantified. Results are given as the mean ± SEM from three independent assays and normalized to the CCL5/RANTES intensity in BFA-treated WT astrocytes. Approximately 1000 cells were quantified in each group. ^ap < 0.001 compared with the DMSO-treated control in WT astrocytes; ^bp < 0.001 compared with the BFA-treated WT astrocytes by Student's t test.

Figure 8.

Abnormal accumulation of astrocytic CCL5/RANTES in the brains of HD mice. A–C, Immunofluorescence staining of CCL5/RANTES (green) and GFAP (red) in brains of R6/2 mice (12 weeks of age; A), Hdh^(CAG)150 mice (16 months of age; B), and WT mice (12 weeks of age; C) was conducted. Nuclei were stained with DAPI (blue). Scale bar, 20 μm.

Figure 9.

Abnormal accumulation of astrocytic CCL5/RANTES in brains of HD patients. A, B, The frontal cortex (A) and caudate nucleus (B) of HD patients and age-matched controls were analyzed. Immunofluorescence staining of CCL5/RANTES (red) and GFAP (green) was conducted. Nuclei were stained with DAPI (blue). Significant amounts of CCL5/RANTES were found in astrocytes of HD patients but not in astrocytes of the age-matched controls. Scale bars, 20 μm.