Effect of 6,633 missense and nonsense variants on MC4R signaling functions.
A. Heatmaps showing the functional effects (z-scores) for nearly all possible amino acid substitutions (6,633 of 6,640) on MC4R activity for two GPCR signaling functions (Gs and Gq) under a variety of conditions. Heatmaps showing results (both z-score and log2[fold change of variant activity over wild-type]) for all experimental conditions are shown in Supplementary Figs. 3-4. The results of the Gs assay with low α-MSH stimulation are highlighted on the left (and in Panels B-F). TM: transmembrane domain; GoF: gain-of-function; LoF: loss-of-function; WT: wild-type activity B. A modified snake plot (adapted from https://gpcrdb.org/) showing the sensitivity of each MC4R residue to mutation (defined as the mean log2[fold change variant activity over wild-type] divided by sqrt(sum(standard error^2)) after excluding nonsense variants). C. Z-scores for each variant (point), broken out by variant type and clinical (ClinVar) classification. Blue indicates statistically significant LoF, brown is significant GoF (significance threshold: FDR < 1%). VUS: variant of uncertain significance. D. Functional effect (Log2[fold change of variant activity over wild-type], x-axis) for all human variants relative to the frequency of the allele in the in the human population (y-axis, gnomAD global population (Chen et al., 2024)). E. DMS results for human MC4R variants (y-axis) relative to previous functional classifications in the literature (x-axis, from (Huang et al., 2017)) F. DMS results (z-scores, x-axis) compared to change in α-MSH potency (relative to WT) of 25 MC4R variants made to the orthosteric binding site for α-MSH (y-axis, cAMP accumulation assay results from (Zhang et al., 2021)). G. Fractions of MC4R variants that result in LoF, GoF, or WT activity for eight unique experimental conditions.