Table 3.

Estimated pK_a shifts due to breaking of the salt-bridge Lys300–Asp163

Simulationa	Asp133	Asp163	Asp164	Lys300
S1	−0.4 ± 0.7	0.8 ± 0.7	−0.1 ± 0.7	−0.5 ± 0.8
S2	−1.5 ± 1.8	0.6 ± 1.3	−0.4 ± 1.6	−1.9 ± 1.6
S4	−1.0 ± 1.1	1.9 ± 1.2	0.2 ± 1.4	−2.1 ± 1.2
S3	−1.4 ± 0.8	3.0 ± 1.5	0.6 ± 1.6	−3.7 ± 1.4
S5	−0.1 ± 0.7	0.6 ± 0.9	0.5 ± 0.7	−1.1 ± 0.8
S6 + S8	0.4 ± 2.1	2.1 ± 2.1	−0.0 ± 1.6	−2.8 ± 1.5
S7	−0.1 ± 0.6	0.3 ± 0.9	1.3 ± 1.2	−1.2 ± 0.8
All data combinedb	−1.5 ± 1.9	1.5 ± 1.8	−1.3 ± 2.4	−2.5 ± 1.4
All data combined, Na+ excluded from calculationc	−1.3 ± 1.5	2.6 ± 1.9	−1.0 ± 2.1	−2.4 ± 1.4
All data combined, salt-bridge cutoff 3.5 Åd	−1.4 ± 1.9	1.4 ± 1.8	−1.3 ± 2.4	−2.5 ± 1.5
All data combined, εsurface = 80e	−1.4 ± 2.0	1.5 ± 1.9	−1.0 ± 2.5	−2.5 ± 1.5

^aFor simulation names, see Table 1. Data were aggregated for all repeats and protomers A and B. Protein conformations were sampled every 1 ns. Na⁺ ions were included within 6 Å of the protein, and the salt bridge was considered broken if the distance was >4 Å (except where noted).

^bDistributions of pK_a and ΔpK_a were recomputed from all trajectory frames of all simulations (i.e., simulated with differing protonation states), assuming that each frame was an independent sample of the protein conformation. This can lead to distributions that are broader than those for individual simulation sets with means that are not just a simple average of individual means.

^cNa⁺ ions were not included in the PROPKA 3.1 calculation.

^dSalt bridge was considered broken if the distance was ≥3.5 Å.

^eThe dielectric constant near the protein surface (corresponding to the solvent-exposed residues) is 160 in the standard PROPKA 3.1 parametrization. A value of 80 was tested, as suggested for calculations on NMR ensembles (Søndergaard et al., 2011).