Structural comparisons of Nap1 ₂ •H2A-H2B•Kap114•Ran ^GTP with other structures. (A) The H2A-H2B heterodimers of Nap1₂•H2A-H2B•Kap114•Ran^GTP (purple), Nap1₂•H2A-H2B•Kap114 (pink; left) and Ran^GTP•Kap114•H2A-H2B (yellow; right, PDB: 8F1E) were aligned. The Cα r.m.s.d. values of different molecules, calculated in PyMOL without realignment, are reported. (B) Left to right: Interactions of H2A-H2B with HEAT repeats 2–5 of Kap114 or lack thereof, and the persistent interactions with h18-19loop of Kap114 and h17-h19 in Nap1₂•H2A-H2B•Kap114 and Kap114•H2A-H2B (PDB: 8F0X). (C) Nap1 mutations at its interfaces with Kap114 or H2A-H2B in Nap1₂•H2A-H2B•Kap114 do not affect H2A-H2B binding as seen in fluorescence polarization assay using 10 nM Nap1₂ FL labeled with XFD488 (Nap1 FL⁴⁸⁸). Data points are averages ± SD. of triplicate measurement. Top: WT Nap1 tiration. The lines show data fitted with one- or two-site binding and residuals are plotted below. Dissociation constants are recorded with the 95% confidence interval obtained by error-surface projection method in brackets. The data is better fitted with two-site binding, which is consistent with previous works by Ohtomo et al. that reported human Nap1₂ binding two copies of H2A-H2B, one bound to the C-terminal acidic tails and one to the core. Bottom: Nap1 mutants β_ERQ and α5/6_mut. One-site binding model was used for fitting as data could not be fitted confidently with two-site binding. All Nap1 mutants bound H2A-H2B with high affinity in the low nM range.