MATERIALS AND METHODS
Molecular cloning and site-directed mutagenesis. Human Ncb5or contains 521 amino acid residues that comprise the N-terminal region and the b5, b5R and CS domains as previously defined.8 These segments of the protein contain 50, 113, 111, and 247 residues, respectively, that also include linker sequences in the middle two fragments (Figure 1A ). Nomenclature of the constructs used herein is as follows: N/b5 (Met1 through Lys137); N/b5-Δ21 (Gly22 through Lys137); N/b5-Δ34 (Met35 through Lys137); b5 (Lys51 through Lys137); and N-term (Met1 through Leu50). The cDNAs of all Ncb5or constructs (except N-term) and rice RLF (residues 101-218, XP_015647767) were expressed with no epitope tag (pET19b or pET22b). The cDNA of N-term was cloned into the pE-SUMOpro Kan vector (Life Sensors, Malvern, CA), which contains the 6XHis-SUMO gene immediately in front of the multiple cloning site previously described.10Missense point mutants, N/b5-LMAA (Leu34Ala/Met35Ala) and N/b5-W37A (Trp37Ala) (Figure 1B ), were generated using the QuikChange mutagenesis kit (Strategene, La Jolla, CA). All mutagenesis primers were synthesized by Integrated DNA Technology (Coralville, IA). All constructs were confirmed by DNA sequencing (ACGT, Inc., Wheeling, IL).
Protein preparation. Except for N-term, all constructs of human Ncb5or and rice RLF were expressed in E.coli BL21(DE3) or BL21(DE3)pLysS or BL21(DE3)pRARE cells and purified as previously described.8 In combination with size exclusion chromatography on a Superdex 75 10/300 GL column, ion-exchange chromatography was performed for b5 and N/b5-Δ34 on a Q HP column, N/b5 and variants (N/b5-W37A, Nb5-LMAA, and N/b5-Δ21) on a SP HP column, and rice RLF on both Q and SP HP columns. Full-length Ncb5or and its variant Ncb5or-Δ50 were prepared as previously described2 , except that 0.5 mM IPTG was used for induction in TB media at 15oC. All purification steps were performed at 4oC. SDS-PAGE was used to estimate the purity of each protein used in spectroscopic analyses, kinetics studies and crystallization screening (all greater than 95%). All purified proteins were flash frozen in liquid nitrogen and stored as aliquots at -80oC until use, except for samples prepared for crystallization screening which were stored at 4oC. The size of each polypeptide product was confirmed by mass spectrometry. Expected sizes are: N/b5, N/b5-W37A, N/b5-LMAA, ~15.6 kD; N/b5-Δ21 and RLF, ~ 13.6 kD; N/b5-Δ34, 11.9 kD; and Ncb5or-b5, 10.2 kD. The ratios of A413/A280 for purified heme-containing constructs were as follows: b5 ≥ 4, N/b5 and variants, RLF ≥ 3.6, Ncb5or ~1.0, and Ncb5or-Δ50 ~1.1. The FAD content of Ncb5or-b5R was determined by A461 and used to represent enzyme concentrations as previously described.8 Final protein yields (mg/L): 8 (b5), 5 (N/b5 and variants), 2 (b5R), ≤1 (Ncb5or, Ncb5or-Δ50, RLF). Solubility in low-salt buffer: N/b5, RLF > b5 > Ncb5or > Ncb5or-Δ50 > Ncb5or-b5R. Protease inhibitor (Sigma, St. Louis, MO) was used during protein preparation to prevent proteolytic cleavage at the b5-CS border of Ncb5or and Ncb5or-Δ50. A SUMOpro Expression System was used to prepare N-term.10Briefly, a 6XHis-SUMO-N-term fusion protein was expressed in BL21(DE3) cells and purified by Ni-NTA affinity resin (Qiagen, Valencia, CA). It was then cleaved by a 6XHis-SUMO protease11 to release 6XHis-SUMO tag, both of which were removed by nickel resin. The purity and size of the N-term peptide were confirmed by SDS-PAGE and mass spectrometry (data not shown).
Spectroscopy. UV/visible spectra were obtained on a Cary 100 Bio spectrophotometer (Varian). The concentrations of oxidized heme in all constructs were determined with ε413 = 130 mM-1cm-1 as previously described.2Circular dichroism (CD) analyses were performed on a JASCO-815 spectropolarimeter using protein concentrations of 4-10 µM. Far-UV (190-250 nm) spectra were recorded using a 1 mm cuvette, whereas a 1 cm cuvette was used for near-UV (250-350 nm) and visible (350-550 nm) CD spectra. All CD spectra are reported in units of molar ellipticity ([θ], deg·cm2·dmol-1).
RLF crystallization. A purified RLF construct (Lys101-Glu218) was concentrated to 44.6 mg/mL in 20 mM Tris pH 8.0 for crystallization screening. All crystallization experiments were set up using an NT8 drop setting robot (Formulatrix Inc.) and UVXPO MRC (Molecular Dimensions) sitting drop vapor diffusion plates at 18oC. 100 nL of protein and 100 nL of crystallization solution were dispensed and equilibrated against 50 µ L of the latter. Crystals approximately 300-500µ m long that displayed a needle morphology were observed within one day from the JCSG+ screen (Molecular Dimensions) condition A6 (20% (w/v) PEG 1000, 100 mM phosphate-citrate pH 4.2, 200 mM lithium sulfate). A cryoprotectant solution composed of 80% crystallization solution and 20% (w/v) glycerol was dispensed (2 µ L) onto the drop, crystals were harvested with a cryoloop immediately and stored in liquid nitrogen. X-ray diffraction data were collected at the Advanced Photon Source beamline 17-ID using a Dectris Pilatus 6M pixel array detector.
Data collection and structure refinement. Intensities were integrated using XDS12, 13 via Autoproc14 and the Laue class analysis and data scaling were performed with Aimless15 which indicated that the crystals belong to the 2/m Laue class with possible space groups P 2 or P 21. Structure solution was conducted by molecular replacement with Phaser16 using the Ncb5or b5-domain structure (PDB 3LF58 ) as the search model. The top solution was obtained in the space group P 2 and contained a dimer in the asymmetric unit. Initial refinement with Refmac17 yieldedR /R free = 42.1%/42.5% and the model was improved by automated building with Arp/wARP.18Further refinement and manual model building were conducted with Phenix19, 20 and Coot21 respectively. TLS parameters 22, 23 were refined in the later stages to model anisotropic atomic displacement parameters. Disordered side chains were truncated to the point for which electron density could be observed. Structure validation was conducted with Molprobity24 and figures were prepared using the CCP4MG package25 . Polder omit electron density maps were calculated using Phenix.26Structure superposition was carried out with GESAMT.27Crystallographic data are described in Table S1 , and the best data set was at 1.55 A resolution.