Fig 2: Expression of AMPA receptor subunits GluA1 and GluA2 in immature cortical and cerebellar astrocytes contacted by explant axons. (A–C) Co-immunostainings for GFAP (green) and GluA1 (red) in immature cortical astrocytes (A) and in cerebellar astrocytes (B,C). Scale bars: 10 µm. (D,E) Co-immunostainings for GFAP (green) and GluA2 (red) in immature cortical astrocytes (D) and in cerebellar astrocytes (E). (F,G) Quantification of GluA1 (F) and GluA2 (G) positive astrocytes contacted by explant axons.

Fig 3: Differential scaling of synaptic AMPA receptor pools at excitatory synapses during homeostatic scaling. A, E, Histograms were constructed to compare the control and TTX dataset for the average intensity of GluA2 segmented within the Shank2 puncta for global synaptic GluA2 using the C-terminal antibody (A) and surface gluA2 using the N-terminal antibody (E); n = 1893 and 1781 synapses for global and 2259 and 2411 for surface GluA2, for control and TTX dataset, respectively, p < 10-6, KS test. B, F, A total of 1000 random puncta were chosen from both the control and TTX dataset, rank ordered, and plotted to provide the scaling equation for global gluA2 (B; y = 0.95x + 1459) and surface gluA2 (F; y = 1.04x + 0.5). C, G, The random sample of the TTX dataset was scaled down using multiple prospective scaling factors and the factor providing the maximum p value between the control and the scaled-TTX dataset was chosen as the multiplicative scaling factor for global gluA2 (C; 1.19, p = 0.084, KS test) and surface gluA2 (G; 1.08, p = 0.680, KS test). D, H, The cumulative distribution was plotted between the control, TTX, and the scaled-TTX (obtained with the help of the scaling factor) for the global (D) and surface (H) GluA2 levels, indicating no significant difference between the control and the scaled-TTX dataset. A gallery of immunocytochemical images for global and surface distribution of GluA2 is presented in Extended Data Figure 3-1.

Fig 4: Differential scaling of synaptic AMPA receptor pools within functional zones of an excitatory synapses during homeostatic scaling. A, B, Confocal and STED imaging were performed for Shank2 (A) and GluA2 (B). A, An adaptive threshold mask was created for Shank2 observed on the synapse (red + blue), perisynapse (red), and PSD (blue) image, and the puncta regions were used to determine different subsynaptic compartments. Analysis scheme for spatial differentiation of functional zones of the synapse using confocal and STED imaging is presented in Extended Data Figure 5-1. B, GluA2 STED image overlaid with red outline indicating the synapse, blue outline indicating the PSD and the region enclosed between the red and blue regions indicating the perisynaptic compartment. 1–6, The insets display different zoomed regions of the GluA2 STED image. Scale bars: 4.5 μm (A, B) and 1 μm (1–6). C–E, Histograms constructed for the average intensity of GluA2 segmented within the Shank2 confocal (C), Shank2 STED (D), and the area between the Shank2 confocal and STED puncta or the perisynaptic area (E) between the control and the TTX dataset showed significant scaling of GluA2 within various subsynaptic compartments (n = 2217 and 2401 for GluA2 in the synapse, p < 10−6, 2694 and 2438 for GluA2 in the PSD, p < 10−4 and 2276 and 2353 for GluA2 in the perisynapse, p < 10−6, for control and TTX dataset, respectively, KS test). F–H, A total of 1000 random puncta were chosen from each dataset and the control set was plotted against TTX dataset to provide a scaling equation for the GluA2 scaling within the synapse (F; y = 1.08x + 0.84), the PSD [G; y = 0.23x + 16.00 (green) and y = 0.83x + 4.62 (blue)], and the perisynapse (H; y = 1.02x + 0.82). I–K, Representative scaling factors providing the maximum p value between the control and the scaled-TTX dataset was chosen as the multiplicative scaling factor for GluA2 within the synapse, (I) in the PSD (J) and in the perisynapse (K). L–N, Cumulative frequency distribution between the control, TTX, and scaled-TTX dataset showed no significant difference between the control and the scaled-TTX dataset for GluA2 within the synapse (L), PSD (M), and the perisynapse (N). Multiplicative scaling of Shank2 at the PSD is presented in Extended Data Figure 5-2.

Fig 5: Effects of LRRTM2 expression on AMPA receptor subunit GluA1 and GluA2 expression in immature cortical and cerebellar astrocytes contacted by explant axons. (A,B) Fluorescence images of transfected astrocytes (EGFP (control) and LRRTM2 + EGFP, green) and immunostainings for GluA1 (red) in immature cortical astrocytes (A1,A2) and in cerebellar astrocytes (B1,B2). Scale bars: 10 µm. (C) Quantification of the fraction of GluA1 positive astrocytes. (D,E) Fluorescence images of transfected astrocytes (EGFP (control) and LRRTM2 + EGFP, green) and immunostainings for GluA2 (red) in immature cortical astrocytes (D1,D2) and in cerebellar astrocytes (E1,E2). (F) Quantification of the fraction of GluA2 positive astrocytes.