Fig. 2: Co-repressor association is a key determinant of heterogeneity of mutant impact.
a,b, We used mammalian two-hybrid assays to assess the effect of mutations in the ligand-binding domain of LXRα on co-activator (a) and co-repressor association (b). HEK293 cells were co-transfected with plasmids expressing LXRa-VP16, either SRC1-GAL4 (a) or NCOR1-GAL4 (b) fusion constructs with a UAS-TK luciferase reporter construct. At 4 h after transfection, cells were treated with the indicated concentration of T0901317 (T09) for 20 h before luminescence was assayed as an index of VP16 and GAL4-conjugate association. Mutants were analysed in three batches each consisting of three independent experiments and are presented as normalized WT luciferase activity (%) in each batch. The mutations are separated by their classification for clarity of presentation and, where possible, are presented alongside the corresponding WT control according to experimental batch. The DN and LOF groups had mutants present in two and three batches. Therefore, the WT control values presented in these panels are the averages of six independent experiments for the dominant negative group and nine independent experiments for the LOF group. The curves represent lines of best fit generated in a three-parameter model conducted in Prism (GraphPad). For the purposes of curve fitting and plotting on a log axis, the untreated condition is plotted as an order of magnitude less than the lowest dose of agonist used. Formal hypothesis testing of mutation class on co-repressor/co-activator association versus WT was conducted using a mixed-effects model with a post hoc Dunnet’s test implemented in R (Methods). P values from these analyses are presented in Supplementary Table 3. Symbols and error bars represent mean ± s.e.m.
