Fig. 2: The LOAD-risk SNP rs10792832 alters PU.1 binding and PICALM expression in iMGs.

a, Schematic of CRISPR–Cas9 editing of the ASoC SNP rs10792832 in iPS cells (two donor lines CD04 and CD09), iMG differentiation and cellular phenotypic assays. The diagram was created using BioRender. b,c, Representative western blot image (b) and quantification (c) of the risk allele reduced PICALM protein expression. Protein quantity was normalized to β-actin. Gel source data are provided in Supplementary Fig. 1. Each data point represents the measurement of cell lysates of 12 wells of cultures from 1 experiment; for each donor line (CD04 and CD09) and each condition (risk and non-risk), data are from 2 clones, collected from 4 independent experiments for clone 1 and 3 independent experiments for clone 2 (that is, n = 7). A linear mixed model (LMM) was used to test the fixed effect of the risk allele, with the experimental round and clone identity used as nested random factors; two-sided test, nominal P values are shown; *P < 0.05, **P < 0.01, ***P < 0.001. Data are mean ± s.e.m. d, The LOAD-risk allele of rs10792832 is predicted to disrupt the PU.1-binding motif (encoded by SPI1; MA0080.2). e, Heterozygous iMGs (A/G) show a higher Sanger-sequencing peak of allele A (non-risk) than G (risk) for the ChIP-assay product of PU.1 binding (top), as opposed to the equal allelic peak height for genomic DNA (bottom).