Skip to main content
. 2021 May 10;12:2578. doi: 10.1038/s41467-021-22590-6

Fig. 6. Interface, disease-associated and phosphomimetic mutations modulate MyD88 signaling and assembly.

Fig. 6

a, b Effects of MyD88 mutations on LPS-induced signaling and MyD88 clustering were tested in HEK293 cells expressing TLR4, MD2 and CD14, with MYD88 knocked out and stably transfected with an NF-κB-driven mScarlet-I fluorescent reporter. The cells were transfected with plasmids expressing wild-type or mutant V5-tagged MyD88, or empty vector, and then treated with (black bars) or without (grey bars) LPS (100 ng/mL) overnight, immunostained to detect MyD88-V5 and analysed by flow cytometry. Cells with very low expression of MyD88 were used for analysis to avoid spontaneous signaling (Supplementary Fig. 5b, c). The mean ± range from n = 2 independent experiments is shown. The death-domain mutation G80K, which has previously been shown to prevent MyD88 clustering130, and a TIR domain alone construct provided negative controls. a NF-κB activation measured by the geometric mean fluorescence intensity of the mScarlet-positive population relative to LPS-treated cells expressing wild-type MyD88. The dotted line indicates level of activation in cells with empty vector. b The percentage of cells with clustered MyD88 was determined based on the elevated height-to-area ratio of the MyD88 signal, which is observed when MyD88 clusters2 (Supplementary Fig. 5e). c Wild-type MyD88 and mutants were expressed in a cell-free system with an N-terminal GFP tag and the fluorescent samples were analysed by single-molecule spectroscopy on a home-made confocal microscope. To characterize the propensity of wild-type MyD88 and mutants to form higher-order assemblies, the average brightness values (equation (1)) of the proteins were calculated72. The results show that S209R, S244D, P245H and T281P mutants have higher propensity than wild-type MyD88 to oligomerize. The mean ± SEM of n = 3 or n = 2 (F270E and T281P) experiments using different lysate batches with two technical repeats per experiment is shown. The G80K mutant and the TIR domain were used as negative controls.