Figure 4

Novel candidate endfoot gene expression is associated with temporal cortical tau pathology and dementia status. For each gene (A–K, left), expression is quantified for each brain region in cognitively intact subjects (blue, n = 57) and subjects with dementia (red, n = 50). Candidate endfoot gene expression was increased in HIP of subjects with dementia for AMOT (B, t = 2.606, p = 0.014), and MLC1 (G, t = 2.512, p = 0.018). Reduced expression of GLUD1 (F, t = −2.255, p = 0.036) and NDRG2 (H, t = −2.120, p = 0.048) was observed in the TCX among subjects with dementia. All candidate endfoot genes, with exception of NDRG2, PBXIP1 and SLCA1A3 exhibited significant associations with P-tau levels (right). This included ACSS1 (t = 2.573, p = 0.038), AMOT (t = 3.549, p = 0.005), BMPR1B (t = 2.498, p = 0.047), FGFR3 (t = 4.196, p < 0.001), FXYD1 (t = 3.383, p = 0.005), GLUD1 (t = 3.986, p = 0.001), MLC1 (t = 4.115, p < 0.001) and PPAP2B (t = 2.856, p = 0.028). The bars in figures (A–K, left) represent the mean ± the standard deviation (S.D.). The line in figures (A–K, right) represent the least squares (ordinary) fit for all data points that show a significant statistical association. All statistical associations between gene expression and dementia status assessed using logistic regression. All statistical associations between gene expression and pathology indicators were evaluated using OLS regression.