Fig 1.
Localization and function of basement membrane and non-basement membrane extracellular matrix (ECM) molecules.
Detailed view (A) shows basement membrane ECM molecules like laminin, entactin/nidogen-1 and collagen. Laminin interacts within the dystroglycan-dystrophin-dystrobrevin complex to ensure anchoring and stabilization of the aquaporin 4 (AQP4) water channel as well as the inwardly rectifying potassium channel Kir 4.1 in the perivascular basal lamina [32, 33]. Detailed view (B) illustrates the participation of non-basement membrane ECM molecules like brevican and tenascin-R in the tripartite synapse. The tripartite synapse consists of a presynaptic and postsynaptic membrane as well as a closely apposed astrocyte process [44]. By connection to further ECM components (hyaluronan, aggrecan, neurocan, vesican) tenascin-R and brevican form perineuronal nets (PNN). Anchored via the cell-surface glycoprotein CD44, PNNs are presumed to play a role in synapse stability and plasticity [44].
Table 1.
Primary antibodies, pretreatment and dilutions used for immunohistochemistry.
Fig 2.
Astrocyte depletion between 56–77 days post infection (dpi) resulted in a deterioration of clinical signs starting at 71 dpi compared to TMEV infected, NaCl treated, GFAP-transgenic (GSTP) mice (purple asterisks), TMEV infected, ganciclovir treated wildtype (WSTG) mice (green asteriks) and TMEV infected, natrium chloride (NaCl) treated wildtype (WSTP) mice (blue asterisks, A). At 77 dpi TMEV infected, ganciclovir treated, GFAP-transgenic mice (GSTG) mice showed a significant detetioration of rotarod performance compared to the WSTG and WSTP controls (green and blue asterisks, B). Clinical and rotarod data are shown as mean and standard error of mean. Significant differences between GSTG and the respective control groups, as obtained by Kruskal–Wallis test, followed by Mann–Whitney U post hoc tests were indicated by *, p < 0.05. Part of these data were previously published [4].
Fig 3.
Impact of astrocyte depletion upon the deposition of extracellular matrix (ECM).
Quantification of ECM components in the thoracic spinal cord white matter was performed using azan (A-C) and picrosirius red (D-F) staining. Accumulation of ECM components was detected intralesionally within the ventral part of the thoracic spinal cord white matter. At 77 dpi TMEV infected, ganciclovir treated, GFAP-transgenic mice (GSTG) mice (A, D) showed a significantly reduced azan and picrosirius red positive area, respectively, compared to TMEV infected, natrium chloride (NaCl) treated wildtype (WSTP) animals (B, C, E, F). Inserts visualize in more detail the intralesional accumulation of azan (B) and picrosirius red (E) labeled ECM molecules in WSTP animals compared to astrocyte depleted mice (A, D). For each antibody, one cross section of the thoracic spinal cord was evaluated per animal. Data are shown as scatter dot plots. The horizontal bar indicates the mean. Significant differences between GSTG and the control groups obtained by Kruskal-Wallis test followed by Mann–Whitney U post hoc tests are indicated by *, p ≤ 0.05. Bars represent 100 μm in overviews and 50 μm in the inserts.
Fig 4.
Impact of astrocyte depletion upon basement membrane extracellular matrix (ECM) components.
Identification and quantification of basement ECM molecules was performed using immunohistochemistry identifying laminin (A-C) and entactin/nidogen-1 (D-F). Statistical analysis revealed a significant reduction of laminin and entactin/nidogen-1 in the lesioned ventral part of the thoracic spinal cord white matter in TMEV infected, ganciclovir treated, GFAP-transgenic (GSTG) mice at 77 dpi (A, D) compared to the non-astrocyte depleted control groups (B, E). Inserts visualize in more detail the increased intralesional accumulation of laminin (B) and entactin/nidogen-1 (E) in TMEV infected, natrium chloride (NaCl) treated wildtype (WSTP) animals compared to astrocyte depleted mice (A, D). For each antibody, one cross section of the thoracic spinal cord was evaluated per animal. Data are shown as scatter dot plots. The horizontal bar indicates the mean. Significant differences between GSTG and the control groups obtained by Kruskal-Wallis test followed by Mann–Whitney U post hoc tests are indicated by *, p ≤ 0.05. Bars represent 100 μm in overviews and 50 μm in the inserts.
Fig 5.
Effects of astrocyte depletion upon Kir 4.1 potassium channels.
Immunohistochemistry targeting Kir 4.1 potassium channels revealed a significant reduction of Kir 4.1 positive area (asterisks, A) in the thoracic spinal cord of TMEV infected, ganciclovir treated, GFAP-transgenic (GSTG) mice compared to control animals (B, C). Inserts visualize in more detail the intralesional reduction of Kir 4.1 in astrocyte depleted animals (A) compared to the WSTP control group (B). For each antibody, one cross section of the thoracic spinal cord was evaluated per animal. Data are shown as scatter dot plots. The horizontal bar indicates the mean. Significant differences between GSTG and the control groups obtained by Kruskal-Wallis test followed by Mann–Whitney U post hoc tests are indicated by *, p ≤ 0.05. Bars represent 100 μm in overviews and 50 μm in the insert.
Fig 6.
Impact of astrocyte depletion upon non-basement membrane extracellular matrix (ECM) components.
Identification and quantification of basement ECM accumulation in the thoracic spinal cord white matter was performed using immunohistochemistry identifying collagen I (A-C), decorin (D-F), tenascin-R (G-I), and brevican (J-L). Statistical analysis revealed a significant reduction of collagen I in the ventral part of the thoracic spinal cord white matter in TMEV infected, ganciclovir treated, GFAP-transgenic (GSTG) mice (A) compared to TMEV infected, NaCl treated, GFAP-transgenic (GSTP) controls (C). Immunohistochemistry detecting decorin showed a significant decrease of decorin positive area in astrocyte depleted animals compared to their controls (F). Inserts visualize in more detail the increased intralesional accumulation of collagen I (B) and decorin (E) in WSTP control animals compared to astrocyte depleted mice (A, D). Furthermore, tenascin-R was significantly reduced in GSTG mice compared to TMEV infected, NaCl treated, wildtype (WSTP) controls (I). Loss of tenascin-R in GSTG animals was intralesionally observed in the thoracic spinal cord white matter as indicated by the arrow in G, whereas control animals showed a diffuse circumferential expression of tenascin-R as visualized in H and in more detail in the inserts (G, H). However, immunohistochemistry targeting brevican did not reveal a significant difference between GSTG mice (J) and the control animals (K, L), as shown also in higher magnification in the inserts (J, K). For each antibody, one cross section of the thoracic spinal cord was evaluated per animal. Data are shown as scatter dot plots. The horizontal bar indicates the mean. Significant differences between GSTG and the control groups obtained by Kruskal-Wallis test followed by Mann–Whitney U post hoc tests are indicated by *, p ≤ 0.05. Bars represent 100 μm in overviews and 50 μm in the inserts.
Fig 7.
CD44 expression during astrocyte depletion.
Immunohistochemistry targeting the cell-surface glycoprotein CD44 was used to evaluate the anchoring of extracellular matrix to astrocytes. Statistical analysis revealed a significant reduction of CD44 positive area in TMEV infected, ganciclovir treated, GFAP-transgenic (GSTG) mice compared to all control groups (A-C). A reduction of the CD44 labeled area in the ventral spinal cord white matter of GSTG animals (asterisk, A) compared to a diffuse and circumferential immunolabelling in control animals (B, C) was detected. For each antibody, one cross section of the thoracic spinal cord was evaluated per animal. Data are shown as scatter dot plots. The horizontal bar indicates the mean. Significant differences between GSTG and the control groups obtained by Kruskal-Wallis test followed by Mann–Whitney U post hoc tests are indicated by *, p ≤ 0.05. Bars represent 100 μm.