Jung et al. 10.1073/pnas.0500722102. |
Fig. 4. The initial electron density (contoured at 1.0 s) of FMN illustrates the high quality of the experimental phases. The final model for FMN is shown.
Fig. 5. Reverse-phase HPLC analysis of flavin cofactors. (A) Reference sample containing 50 mM riboflavin, FMN, and FAD, respectively. (B and C) BlrB sample before (B) and after (C) reconstitution with FAD. (D) Cofactor content of BlrB crystals. The BlrB crystals contain bound FAD.
Fig. 6. Constant time 13C-1H HSQC spectra of BlrB recorded under dark (black peaks) and signaling active state (red peaks) conditions.
Table 1. Reverse-phase HPLC protocol that enables separation of flavin-derived cofactors
Time, min | Flow rate, ml/min | A, % | B, % |
0 | 0.63 | 100 | 0 |
20 | 0.63 | 60 | 40 |
25 | 0.8 | 60 | 40 |
43 | 0.63 | 100 | 0 |
44 | 0 | 100 | 0 |
A, 50 mM NH4Ac, pH 6.0; B, 70% acetonitrile in A.
Supporting Materials
Cloning, Expression, and Purification of BlrB140.
DNA encoding BlrB (residues 1–140) was subcloned into the expression vector pHis-Gb1-Parallel1, a derivative of the pHis-Parallel1 vector (1, 2). Escherichia coli BL21 (DE3) cells containing this vector were grown in LB media at 37°C to a density A600 of 0.6. The temperature was then lowered to 18°C. Cultures were induced with 0.7 mM isopropyl-b-thiogalactoside in the dark when an A600 of 0.8 to 0.9 was reached. After 14 h, expression was stopped by pelletting and resuspending the cells in 50 mM NaPi buffer, pH 8.0. Cells were lysed with a microfluidizer (Microfluidics). The lysate was clarified by centrifugation at 15,000 ´ g for 45 min. The following purification steps were carried out in the dark or under red light to keep the light-sensitive protein in its receptor state. The resulting supernatant was loaded onto a Ni2+-NTA Superflow column (Amersham Pharmacia), allowing the rapid affinity purification of His-Gb1-tagged BlrB fusions by washing with 30 and 50 mM imidazole, respectively, and eluting with 250 mM imidazole. After dialysis of BlrB containing fractions against 25 mM Tris·HCl, pH 8.0/200 mM NaCl/5 mM DTT/5% glycerol, the His-Gb1 tag was cleaved by adding 0.5 mg H6-TEV protease per 40 mg of fusion protein. Proteolysis proceeded overnight at room temperature. To remove free His-Gb1 tag, uncut protein, and the H6-TEV protease, a second Ni2+-NTA chromatography step was performed. The cleaved protein was then concentrated and subjected to a final gel filtration step (Superdex 200, Amersham Pharmacia). Fractions containing pure BlrB were dialysed against storage buffer (25 mM Tris·HCl, pH 8.0/40 mM NaCl/10 mM KCl/2 mM EDTA/5 mM DTT/5% glycerol).Cofactor Analysis and Reconstitution.
Because flavin adenine dinucleotide (FAD) was observed to be unstable even when bound to a blue light sensing using FAD (BLUF) domain, the cofactor content of BlrB was controlled carefully. Therefore, we developed a chromatographic protocol (Table 1) that allows separation and quantification of noncovalently bound FAD, FMN, and riboflavin by using a C18-HPLC column (Hypersil-ODS, Bischoff Chromatography, Leonberg, Germany) connected to a Waters HPLC system.Elution was followed by absorption at 370 nm. Typical retention times for FAD, FMN, and riboflavin were 20.9, 22.2, and 24.8 min, respectively (Fig. 5A). To release the flavin cofactors, concentrated protein samples were heat-denatured for 3 min at 95°C and subsequently centrifuged at high speed. The supernatant was subjected to the HPLC assay described. Purified BlrB140 contained about 42% FAD, 26% FMN, and 32% riboflavin (Fig. 5B). To completely reconstitute BlrB in its physiologically active form, the protein was incubated with a 50-fold excess of free FAD for 4 h at room temperature. Free nucleotides were removed by gel filtration, and the final FAD content was checked and shown to be nearly 100% (Fig. 5C).
To analyze BlrB crystals for their cofactor content, crystals were collected from up to five hanging drops. They were washed five times with mother liquor to remove free protein and flavin molecules and subsequently dissolved in 50 mM Tris·HCl, pH 8.0. After heat treatment, the protein solution was centrifuged. HPLC analysis of the resulting supernatant showed that »60% of the protein molecules within BlrB crystals bind FAD. As a consequence of the cofactor’s instability, the remaining molecules contain a flavin derivative similar to FMN (Fig. 5D).
NMR Spectroscopy.
NMR spectra were obtained from samples containing 0.1 mM BlrB (unlabeled for 1D spectra, Fig. 1B; uniformly 13C-labeled for HSQC spectra, Fig. 6) in the same buffer used for UV-visible spectroscopy. Spectra were recorded at 25°C on a Varian Inova 500 MHz spectrometer equipped with a cryogenically cooled, triple resonance, pulsed field gradient probe. To obtain NMR spectra of the signaling active state, we generated blue light from a 5 W Coherent Inova-90C argon ion laser running in single wavelength mode at 457.2 nm, directing this light into the NMR sample through quartz fiber optics (2). The power level measured at the sample end of this fiber was 120 mW. This measurement was verified before every experiment. For 1D 1H spectra, 512 scans were acquired with a 1.1-s recycle delay by using a 200-ms laser pulse during the recycle delay between scans to obtain spectra of the signaling active state. For 2D constant time 13C-1H HSQC spectra, 512 and 128 complex points were acquired in the 1H and 13C dimensions, respectively. Spectra of the putative signaling active state were obtained by using the same recycle delay and laser pulse widths as with the 1D methods.Quantum Chemical Calculations.
Lumiflavin in the oxidized state (F) and the corresponding protonated cation (FH+) were considered as models for the chromophore in the dark state and the photoproduct, respectively. The structures of the molecules in the ground singlet state (S0) were optimized at the MP2/6-31G(d) level of theory. The lumiflavin structures in the first exited open-shell singlet state (S1) were optimized with two-configuration self-consistent field wave function and the 6-31G(d) basis set. The structural parameters and charge distribution were compared for the lumiflavin in the ground and first exited singlet states. For the structures corresponding to the MP2/6-31G(d) minima, the energy of first three singlet states (S0, S1, S2) were calculated with the CASSCF(8,8)/6-31G(d) method with the energy averaging over the three states considered. The complete active space calculations were carried out for the active space including the four highest occupied and four lowest unoccupied molecular orbitals of lumiflavin and the cations. Energies for the two low energy electron transitions were calculated as DE1 = E(S0) – E(S1) and DE2 = E(S0) – E(S2). The energies calculated for the first two low transitions of lumiflavin were 3.93 and 5.06 eV, which are in reasonable agreement with the experimental data (2.78 and 3.35 eV, respectively). To estimate the basicity of the different possible proton acceptor atoms of lumiflavin, namely N1, N5, and O2, the deprotonation energies, EDP, for the ground and first excited singlet states were determined according toEDP(S0) = E(S0)(F) – E(S0)(FH+), or EDP(S1) = E(S1)(F) – E(S1)(FH+).
The pc gamess software (version 6.2) was used for all calculations (3), and the program molekel (4) was used for the analysis of the results.
Sequence Space Analysis.
Sequence space analysis of the BLUF domain family was performed by using an algorithm introduced by Casari et al. (5). A multiple sequence alignment encompassing >50 different prokaryotic and eukaryotic BLUF domains was manually complemented by the amino acid sequence of the BlrB and BlrA BLUF core and used as input information for the analysis. Results were visualized with a set of Java viewers combining sequence space view with more conventional protein displays (http://industry.ebi.ac.uk/SeqSpace/index.html); (Fig. 3 A and B and Table 3).1. Sheffield, P., Garrard, S. & Derewenda, Z. (1999) Protein Expression Purif. 15, 34–39.
2. Harper, S. M., Neil, L. C. & Gardner, K. H. (2003) Science 301, 1541–1544.
3. Schmidt, M. W., Baldridge, K. K., Boatz, J. A., Elbert, S. T., Gordon, M. S., Jensen, J. H., Koseki, S., Matsunaga, N., Nguyen, K. A., Su., S. J., Windus, T. L., et al.. (1993) J. Comput. Chem. 1347-1363.
4. Flükiger, J. P., Lüthi, S. & Portmann, J. (2003) MOLEKEL, Swiss Center for Scientific Computing (Manno, Switzerland), Version 4.2.
5. Casari, G., Sander, C. & Valencia, A. (1995) Nat. Struct. Biol. 2, 171–178.
>6. Vajdos, F. F., Yoo, S., Houseweart, M., Sundquist, W. I. & Hill, C. P. (1997) Protein Sci. 6, 2297–2307.
7. Wilson, A. (1950) Acta Crystallogr. 3, 397–398.