Dustin J. Maly

Variation alters inherent protein properties like thermodynamic stability and catalytic activity, and these intrinsic effects are shaped by the cellular environment. The RAS–MAPK signaling pathway typifies this interplay, with hundreds of proteins acting in concert through regulated states and interactions to transduce signals. To explore how cell context influences variant effects, we measured the abundance and pathway activity of 186,676 variants across 17 MAPK proteins in human cells under diverse conditions, for a total of 503,099 measurements. We discovered 14,412 dominant negative variants, nearly all previously unknown, that disrupt specific domains or interaction interfaces while preserving other functions. MAPK dominant negatives occur at conserved positions and are depleted from the human population relative to other activity-altering variants, representing a distinct constraint on the human genome. Modulating chaperone activity exposed previously hidden effects of genetic variation on protein abundance in human cells, revealing the HSP90 dependence of 67% of missense variants across the pathway. Combining hundreds of experimental and predicted protein interaction structures with variant effect data distinguished functional from non-functional interaction interfaces and revealed how their disruption affected function. We resolved the overlapping KRAS-LZTR1-SOS1 interface into partner-specific and shared binding determinants, identified allosteric networks supporting each interaction, and uncovered distinct mechanisms by which variants promote LZTR1-mediated degradation. Thus, we measure variant effects at pathway scale in human cells, illuminating fundamental MAPK pathway signaling mechanisms and demonstrating that cellular context profoundly shapes the effect of most variants.

Summary: High-resolution measurements of pathway activity make it possible to characterize how signaling responds to changing cellular conditions, to evaluate pathway-targeted inhibitors, and to dissect how sequence variation shapes signaling output. Here, we describe a sequencing-based assay that uses engineered, phospho-state-selective affinity reagents to quantify ERK activation state as a high-resolution, ratiometric readout of RAS/ERK MAPK pathway activity. Profiling ∼1,600 BRAF variants at basal and elevated RAS-GTP levels allowed us to assign activity scores to somatic cancer and germline RASopathy variants and to distinguish residues that stabilize autoinhibition from those required for active dimer formation. We identified a large class of dominant negative variants that suppress signaling below endogenous levels through enhanced RAS engagement and an inability to form productive RAF dimers. Measuring the intracellular potency of the RAF inhibitor LXH254 against ∼1,000 variants in parallel revealed hinge and activation-loop positions conferring resistance and showed that disrupting the BRAF-MEK interface is broadly sensitizing.

Figure S11 shows glucose uptake assays.

Figure S7 shows Western blot analysis confirming the findings with RPPA.

Table S4 shows screening results of BKIDCs tested at 10 µM in LNCaP95 and PC3 cells in the MTS assay.

Table S7 shows pharmacokinetic parameters after a single dose in mouse, rat, dog, and cynomolgus monkey.

Table S3 shows PrP and PP-based BKIDC chemical structures.

Figure S1 shows evaluation of Hexokinase activity using LNCaP crude extracts and recombinant enzyme.

Table S6 shows in vitro AMPK activity.

Figure S3 shows competitive proteomic profiling of 1553-N-Me.