Table I. Structural statistics.

	<SA>a	(SA)rb
A. R.m.s. deviations from restraints	0.064 ± 0.001	0.060
protein
inter-residue sequential (|i – j| = 1) (664)	0.041 ± 0.002	0.036
inter-residue medium range (1 < |i – j| ≤ 5) (504)	0.053 ± 0.003	0.042
inter-residue long range (|i – j| > 5) (212)	0.062 ± 0.006	0.062
intraresidue (616)	0.020 ± 0.002	0.020
intersubunit (174)	0.038 ± 0.005	0.034
DNA
intraresidue (428)	0.108 ± 0.001	0.105
sequential intrastrand (196)	0.058 ± 0.003	0.054
interstrand (24)	0.097 ± 0.006	0.080
protein–DNA (168)	0.092 ± 0.005	0.067
R.m.s. deviations from hydrogen bonding restraints (Å)c protein (138)	0.030 ± 0.004	0.046
DNA (120)	0.013 ± 0.001	0.017
R.m.s. deviations from repulsive restraints (Å) (106)d	0.011 ± 0.009	0.004
R.m.s. deviations from experimental dihedral restraints (°) (708)e	0.338 ± 0.044	0.320
R.m.s. deviations from 3JHNα coupling constraints (Hz)d (72)	1.00 ± 0.05	1.12
R.m.s. deviations from secondary 13C shifts (p.p.m.)d
13Cα (140)	0.98 ± 0.01	0.99
13Cβ (140)	1.42 ± 0.04	1.36
R-factor for residual dipolar couplings (%)f
protein 1DNH (70)	1.46 ± 0.04	2.32
DNA 1DCH (70)	3.78 ± 0.16	4.58
Deviations from idealized covalent geometry
bonds (Å) (1946)	0.008 ± 0.000	0.007
angles (°) (3524)	1.053 ± 0.005	1.146
impropers (°) (979)	1.782 ± 0.026	1.791
B. Measures of structure quality
EL–J (kcal/mol)g	–1136 ± 18	–1153
% residues in most favorable region of Ramachandran mapg	88 ± 2	87
C. Coordinate precision (Å)h
Protein backbone + DNA heavy atoms	0.34 ± 0.06
Protein heavy atoms + DNA heavy atoms	0.54 ± 0.05
Protein backbone	0.37 ± 0.07
Protein heavy atoms	0.65 ± 0.06
DNA heavy atoms	0.31 ± 0.09

^a<SA> is the final set of 35 simulated annealing structures.

^b is the mean structure obtained by averaging the coordinates of the 35 individual SA structures with residues 1–73 of both subunits of the protein and bp –10 to +10 of the DNA best-fitted to each other; ()r is the restrained regularized mean structures obtained by restrained regularization of (Nilges et al., 1988). The total numbers of restraints are given in parentheses. The number of unique restraints is half that since the protein is a symmetric homodimer and the DNA is symmetric palindrome. None of the structures exhibited interproton distance violations >0.5 Å, torsion angle violations >5°, or ³J_HNα coupling constant violations >2 Hz.

^cBackbone hydrogen bonding restraints (two per hydrogen bond) were added during the final stages of refinement according to standard criteria (Clore et al., 1989). Six distance restraints per base pair are used to represent the Watson–Crick base pairs: for GC base pairs, r_N1–N3 = 2.87 ± 0.2 Å, r_H1–N3 = 1.86 ± 0.2 Å, r_O6–N4 = 2.81 ± 0.2 Å, r_N2–O2 = 2.81 ± 0.2 Å, r_O6–N3 = 3.58 ± 0.2 Å and r_N2–N3 = 3.63 ± 0.2 Å; for AT base pairs, r_N1–N3 = 2.92 ± 0.2 Å, r_N1–H3 = 1.87 ± 0.2 Å, r_N6–O4 = 2.89 ± 0.2 Å, r_H2–O2 = 2.94 ± 0.2 Å, r_N1–O4 = 3.69 ± 0.2 Å and r_N1–O2 = 3.67 ± 0.2 Å. The O6–N3 and N2–N3 distance restraints for GC base pairs and the N1–O4 and N1–O2 distance restraints for AT base pairs serve to prevent unduly large shearing of the base pairs.

^dRepulsive distance restraints with a lower bound of 4 Å and no upper bound were introduced in the final stages of the refinement to prevent energetically unfavorable proximity of hydrogen bond donors of the protein (involving Gly1, Arg2, Lys3, Lys4, Arg14 and Lys22) to hydrogen bond donors of the DNA (Omichinski et al., 1997).

^eThere are 240 torsion angle restraints for each protein subunit (71 φ, 71 ψ, 56 χ1, 34 χ2 and 8 χ3) and 114 per strand of DNA (see Materials and methods).

^fThe dipolar coupling R-factor is defined by the ratio of the r.m.s. deviation between observed and calculated values to the expected r.m.s. deviation if the vectors were randomly oriented. The latter is given by {2D_a²[4 + 3R²]/5}^1/2 where D_a is the magnitude of the alignment tensor and R the rhombicity (Clore and Garrett, 1999). The dipolar couplings per strand of DNA are broken down into 11 for the bases and 24 for the sugars.

^gE_L–J is the Lennard–Jones van der Waals energy calculated with the CHARMM PARAM19/20 protein and PARNAH1ER1 DNA parameters and is not included in the target function for simulated annealing or restrained regularization. The overall quality of the protein moiety (residues 1–73) was assessed using the program PROCHECK (Laskowski et al., 1993). There are no φ,ψ angles in the disallowed region of the Ramachandran map. The dihedral angle G-factors for φ/ψ, χ1/χ2, χ1 and χ3/χ4 are –0.02 ± 0.03, 0.02 ± 0.07, –0.19 ± 0.11 and –0.32 ± 0.13, respectively.

^hThe precision of the coordinates is defined as the average atomic r.m.s.d. between the 35 individual simulated annealing structures and the mean coordinates. The values refer to residues 1–73 of both subunits and bp –10 to +10 of the DNA. Residues 74–85 of the protein are disordered in solution.