Astrocytic engagement of the corticostriatal synaptic cleft is disrupted in a mouse model of Huntington’s disease

Significance Astrocytic physiological dysfunction contributes to development of the neurodegenerative phenotype in Huntington’s disease (HD), but the structural correlates to this dysfunction are unclear. Here, we used a combination of viral tracing, phenotype-specific tagging, and ultrastructural modalities to reconstruct HD synapses at the nanometer scale in the neostriata of HD mice. We discovered significant impairment in the astroglial engagement of mature striatal synapses. In light of the known deficiencies in glutamate and potassium uptake by HD astrocytes, these findings suggest the potential for leakage of excitatory synaptic contents during neurotransmission, and hence a structural basis for neuronal hyperexcitability in HD. More broadly, our data suggest that astrocytic structural pathology may causally contribute to those neurodegenerative disorders associated with central hyperexcitability.


Figure S1
Related to Figure 1 MicroCT mapping of EM block region containing the desired region of interest (Related to Figure 1) A-B. Extracted section of striatum after osmium infiltration and resin polymerization. Sides of block ~500µm, thickness ~600 µm. MicroCT scan of resin block leaves internal structure intact and highlights contrast differences between grey and white matter, as well as vascular cavities and induced ablations, the latter as empty (black greyscale values) (A). Comparison of single slice at MSN-astrocyte region of interest (ROI), with highlighted structures relocalized in the microCT scan (B). C. Use of microCT data to determine location depth of ROI, after cross referencing correct location from 2-photon stack. D. Comparison of expected readout from microCT (left) and initial surface sections of serial-blockface SEM data (right), highlighting induced ablations (yellow arrow), myelinated axons (green arrow), and corresponding cell soma (blue and red arrows). Dimensions of measured distances are also shown to match (violet measurements). This confirms expected number of microtome slices remaining to astrocyte-neuron ROI branches, and hence angle estimate at which the diamond blade is cutting the block-face during sample approach. Thus, the microCT data allows us to adjust in real-time the acquisition parameters as the ROI appears in the field of view. Scale bars: 50 µm.  Similarly, thin spine volume did not differ between R6/2 and WT mice (bottom panel; p=0.2).

B. Equivalent surface area (top panel) and volume (bottom) of dendritic spines across
groups, for synaptic interactions with mature spines (p=0.3) and immature spines (p=0.5), respectively.
C. Significantly increased dendritic spine volume was noted, however, for mature mushroom spines in both WT (top panel, *** p<0.001) and R6/2 mice (bottom, *** p<0.001), indicating that mature spines are larger and more complex than thin/immature spines in both WT and R6/2 MSNs.     Figure 1 X-ray datasets of samples were used as a main navigation tool, whereby the higher resolution 2 photon datasets from preceding captures are overlayed onto both fiduciary marks and the microvasculature, to determine the correct x-y-z coordinates of the prospectively defined neuron-astrocyte region-of-interest (ROI), for subsequent serial SEM imaging. Note that the reconstructed MSN details the difference between the 2 photon-segmented and serial EM datasets.

Video 2
Relocalization of EM stack, after X-ray and 2 photon correlated light EM (CLEM) Related to Figure S1 This video presents the EM environment with iteratively higher magnification scans. Multiple EM stack correlations are necessary, since serial block-face SEM is a destructive imaging process, and there is only one opportunity to capture each desired ROI stack. The reconstructed medium spiny neuron branch shown here is the same as that in video 1, with dendritic spines added to show the neuron-glia environments selected for analysis.

Video 3
Reconstructed neuropil volume of striatum and perisynaptic astrocyte Related to Figure 2 This selected subvolume of striatum shows the density of the segmented environment. The camera path is focused on a single synapse, and shows a prototypic example of astrocytic coverage of the perisynaptic region. Multiple views reveal the complex association of astrocytes with both the pre-and post-synaptic components of mature, mushroom-like synapses, the proximal elements of which were used for the analysis strategy described in Figure 2 (see segment with opaque and transparent overlay of astrocyte, at 00:22:00).

Video 4 Identification and characterization of perisynaptic morphologies Related to Figures S2 and S3
We observed a variety of configurations of astrocytic engagement with MSN spines, with a spectrum of astrocytic infiltration patterns observed at each analyzed level of dendritic spine maturity. These examples focus on the prototypic engagement of mature spines in WT mice, in which tight astrocytic sequestration of the perisynaptic space was typically observed.