Figure 4.

Electrostatic contributions to the cooperativity in the BRD4^BD2-VHL system are small and context-dependent. All calculations shown are obtained using MBAR, and results using TI and BAR are shown in Figure S6. (a) Calculations of ΔΔGs over PROTAC linker lengths are shown with the experimental measurements²⁸ (black) converted to our units. Experimental results at 2, 3, and 4 linker beads correspond to MZ4, MZ1, and MZ2 for PROTACs using JQ1 warhead and MZP-61, MZP-54, and MZP-55 for PROTACs using I-BET726 warhead. (b) Waterfall plot breakdown of ΔΔG calculations. At each linker length, bars in each triplet correspond to ΔG^binary (gray), −ΔG^{ternary(other)} (light purple), and −ΔG^{ternary(charges)} (turquoise), and are arranged in a cumulative manner such that the end position marks the resulted ΔΔG (orange). ΔG^{ternary(other)} denotes the free energy change of turning on interaction energy components other than the electrostatics, which only include steric repulsions in this panel. (c) ΔΔG breakdowns at linker length 3 under different force field parametrizations are superimposed for comparison. Reducing the screening effect by 10-fold (charges *) significantly increases ΔG^{ternary(charges)} (cyan), which leads to a very negative ΔΔG. Introducing nonspecific attractions (ϵ_LJ = 0.2 kT) not only reduces ΔG^{ternary(other)} (dark purple) but also doubles ΔG^{ternary(charges)} (steel blue), resulting in a positive ΔΔG.