Fig 2: Validation of the Panx1-Crmp2 interaction in the mouse ventricular zone (VZ). (Ai) Western blot of a time course of Panx1 and Crmp2 expression in the VZ. (ii) Panx1 expression peaked within the early postnatal period (P0–P7), and dramatically decreased with age (P28–P60; p = 0.0365 by one-way analyses of variance (ANOVA); F(4,10) = 3.9; Dunnett’s multiple comparison test post hoc, P28 p = 0.0352, and P60, p = 0.0177; n = 3). Coomassie Blue was employed as a loading control. (iii) Crmp2 expression remained relatively constant across this time course (F(4,10) = 1.3; p = 0.3333, one-way ANOVA; n = 3). (B) Western blot of endogenous Panx1 IPs from mouse brain VZ tissue. Crmp2 co-precipitated specifically with Panx1 from young (P10; i) and adult (P60; ii) mouse VZ. For the complete blots of part (B), see Supplementary Material. This data was previously included in the PhD thesis of LWS (link: https://dspace.library.uvic.ca/handle/1828/6938).

Fig 3: Probenecid treatment results in Crmp2 concentration at neurite tips. (Ai) Representative confocal images from N2a cells treated with vehicle (upper panels) or probenecid (lower panels) and immunostained for Crmp2. Crmp2 signal intensity is indicated in the heat maps. Hoechst 33342 was used as a nuclear counterstain. Scale bars: 20 µm. (ii) Crmp2 signal in probenecid-induced neurites was more abundant in neurite tips compared to the shaft, regardless of neurite length (10–30 µm, p < 0.0001, n = 38 neurites; 30–50 µm, p = 0.0003, n = 40 neurites; >50 µm, p < 0.0001, n = 48 neurites; Wilcoxon signed-rank test against a hypothetical median value of 1.0).

Fig 4: Probenecid treatment decreases Crmp2 co-precipitation with Panx1. (Ai) Western blot of a GFP IP from N2a cells expressing Panx1-EGFP, or EGFP control, treated (16 h) with probenecid or vehicle control. (ii) Probenecid treatment (16 h) significantly decreased the amount of Crmp2 that precipitated with Panx1-EGFP. Crmp2 signal in the IP fraction was normalized to GFP signal intensity from the same lane (Panx1-EGFP [vehicle] vs. Panx1-EGFP [probenecid]: 100.0 ± 0.8 vs. 39.4 ± 19.6, p = 0.0368, n = 3, unpaired t-test). (B) The same assay was performed with a shorter 30 min treatment (i). (ii) Short probenecid treatment (30 min) revealed a ~30% decrease in Crmp2 co-precipitation (Panx1-EGFP [vehicle] vs. Panx1-EGFP [probenecid]: 100.0 ± 11.1 vs. 67.7 ± 19.6, p = 0.0826, n = 3, unpaired t-test). (C) PNGase F treatment was performed to confirm the upper Panx1EGFP immunoreactive band was indeed glycosylated Panx1 and assessed by Western blotting. Note that both immunoreactive bands shifted in response to N-glycosidase F treatment, suggesting that both major bands are mature glycosylated Panx1EGFP of differing molecular weights. (D) Analysis of the input intensities revealed that probenecid treatment did not change the overall expression of (i) Crmp2 (p = 0.2000, Mann-Whitney test, n = 3) or (ii) Panx1 (p = 0.578, unpaired t-test, n = 3). Normalized to Gapdh (not shown), n.s., non-significant.

Fig 5: Pannexin 1 (Panx1) and collapsin response mediator protein 2 (Crmp2) interact. (A) Crmp2 peptides (red) identified previously by immunoprecipitation (IP) coupled to LC-MS/MS in Wicki-Stordeur and Swayne (2013). Peptides were identified specifically in IPs from Panx1-EGFP expressing Neuro-2a (N2a) cells, and not EGFP controls. (B) Western blot of endogenous IPs from N2a cell lysates. (i) Crmp2 precipitates specifically with Panx1 compared with IgG control. (ii) Panx1 precipitates specifically with Crmp2 compared with IgG control. (C) Western blots of an in vitro binding assay using purified Crmp2 and Panx1 C-terminus (Panx1CT). (i) The Panx1CT enriches with immobilized Crmp2-GST compared with GST control. (ii) Crmp2 enriches with Panx1CT-GST compared with GST control. These results are representative of three independent replicates. For the complete blots of parts (B,C), see Supplementary Material. This data was previously included in the PhD thesis of LWS (link: https://dspace.library.uvic.ca/handle/1828/6938).