Fig. 6

Correlation of GOAC proteinuria with clinical data and MN mechanism modeling on the chip. a Table of clinical parameters for proteinuria in MN serum samples used on the GOAC. b hAKPC-P + hGEC chip proteinuria for CTRL1-2 and MN1–6 clinical proteinuria levels suggests a very strong correlation between clinical profile and response in the chip (measured as albumin leakage). R: 0.8901, P < 0.01. c hiPOD + hGEC chip proteinuria for CTRL1–2 and MN1–6 clinical proteinuria levels suggests a weak correlation between clinical profile and response in the chip (measured as albumin leakage). R: 0.3156, not significant. d hpPOD + hGEC chip proteinuria for CTRL1–2 and MN1–6 clinical proteinuria levels suggests a strong correlation between clinical profile and response in the chip (measured as albumin leakage). R: 0.7995, p < 0.05. For all samples, regression analysis was performed. Equation: Polynomial, linear. Blue lines = 95% confidence band; red lines = 95% prediction band. e, f Western blot analysis for C3d (140kDA) and beta actin (40 kDa) in hAKPC-P + hGEC, hiPOD + hGEC, and hpPOD + hGEC chips exposed to healthy or MN serum confirmed increased expression for C3d by all three MN chips. Number of replicates per experimental group: 3 (f). Quantification of C3d expression was performed by measuring pixel density and followed by normalization against beta actin. g, h Western blot analysis for NPHS1 (138kDA) and beta actin (40 kDa) in hAKPC-P + hGEC, hiPOD + hGEC, and hpPOD + hGEC chips exposed to healthy or MN serum confirmed decreased expression for NPHS1 by all three MN chips. Number of replicates per experimental group: 3. h Quantification of NPHS1 expression was performed by measuring pixel density and followed by normalization against beta actin (g). For all samples lack of significant differences was determined by a one-way ANOVA and Student–Newman–Keuls post hoc test. Box plots show the median, the 25th and 75th percentiles, whiskers (median ± 1.5 times interquartile range), and outliers (solid circle). Western blot images were cropped to show the relevant bands and improve clarity. Uncropped and unprocessed scans for western blotting analysis are provided in Supplementary Figs. 13, 14