Electrochemical and structural characterization of Azotobacter vinelandii flavodoxin II

Abstract Azotobacter vinelandii flavodoxin II serves as a physiological reductant of nitrogenase, the enzyme system mediating biological nitrogen fixation. Wildtype A. vinelandii flavodoxin II was electrochemically and crystallographically characterized to better understand the molecular basis for this functional role. The redox properties were monitored on surfactant‐modified basal plane graphite electrodes, with two distinct redox couples measured by cyclic voltammetry corresponding to reduction potentials of −483 ± 1 mV and −187 ± 9 mV (vs. NHE) in 50 mM potassium phosphate, 150 mM NaCl, pH 7.5. These redox potentials were assigned as the semiquinone/hydroquinone couple and the quinone/semiquinone couple, respectively. This study constitutes one of the first applications of surfactant‐modified basal plane graphite electrodes to characterize the redox properties of a flavodoxin, thus providing a novel electrochemical method to study this class of protein. The X‐ray crystal structure of the flavodoxin purified from A. vinelandii was solved at 1.17 Å resolution. With this structure, the native nitrogenase electron transfer proteins have all been structurally characterized. Docking studies indicate that a common binding site surrounding the Fe‐protein [4Fe:4S] cluster mediates complex formation with the redox partners Mo‐Fe protein, ferredoxin I, and flavodoxin II. This model supports a mechanistic hypothesis that electron transfer reactions between the Fe‐protein and its redox partners are mutually exclusive.

the day of purification. Dioxygen was removed from the size exclusion column buffer prior to use in the purification by iterative cycles of vacuum followed by argon filling.
An overnight starter culture of the flavodoxin II overexpression strain supplemented with 100 µg/mL ampicillin was incubated at room temperature with shaking overnight. The starter culture was diluted 1:100 into 1 L of LB medium containing 50 µg/mL ampicillin. The cultures were grown to late exponential phase (OD 600 ~ 0.6-0.8) at 37 °C with shaking. Overexpression of flavodoxin II was induced by addition of IPTG to a final concentration of 0.5 mM in each culture. The cultures were incubated for 16 hours at 37 °C with shaking. The cells were pelleted via centrifugation at 6238 x g for ten minutes. The cell pellets were stored at -80 °C until purification.
All lysis steps were carried out on ice at room temperature. The BL21 (DE3) cell pellets were thawed on ice. The cell pellets were re-suspended in anion exchange buffer A with a homogenizer. The lysis buffer also contained complete protease inhibitor tablets (Roche) at a concentration of one tablet per 50 mL of buffer. The cells were lysed with an Avestin Emulsiflex C5 homogenizer. The cell debris was pelleted by centrifugation in a floor centrifuge at 13,000 x g at 4 °C for 35 minutes.
The cleared cell lysate was loaded onto two 5 mL HiTrapQ HP anion exchange columns connected in tandem using an Akta FPLC (GE Healthcare). The protein was purified by anion exchange chromatography as described previously 19,55 . Following the anion exchange column, the protein was loaded onto a Superdex200 size exclusion column. The protein that eluted from the column was collected, and was concentrated in an Amicon filter centrifuge tube (10,000 molecular weight cutoff) to a final concentration of about 10 µM. This protein was flash frozen in liquid nitrogen and stored under liquid nitrogen.
The concentration of protein was estimated based on the absorption of the sample at 452 nm, using an extinction coefficient of 11,300 M -1 cm -1 55 . The total protein yield was about 0.2 mg of protein per gram of cell paste. The purity of the isolated protein was analyzed on a denaturing polyacrylamide gel. The concentration of protein with bound flavin mononucleotide cofactor was estimated from the ratio of the absorption of the sample at 274 nm to the absorption of the sample at 452 nm. The ratio of Abs 274 /Abs 452 was typically ~7, which was higher than the ratio of 4.6-4.8 reported for the oxidized form of this protein 55 , suggesting the presence of some apo or partial reduced forms of the flavodoxin.

Preparation of Flavodoxin II for Electrochemistry Experiments.
Prior to electrochemistry experiments, sodium dithionite was removed from all protein samples using a PD10 column (GE Healthcare). This procedure was performed in a McCoy anaerobic chamber with a 95 % / 5 % Ar / H 2 atmosphere. The column was equilibrated in five column volumes of electrochemistry buffer (50 mM potassium phosphate, pH 7.5, 150 mM NaCl) with 10 mM sodium dithionite. This step allowed for the reduction of reactive oxygen species prior to introduction of the protein to the column. Then, the column was equilibrated in five column volumes of electrochemistry buffer to remove all sodium dithionite. All buffers were filtered with a 0.2 micron filter, and oxygen was removed from the buffer by iterative cycles of vacuum followed by filling with argon.
The protein was exchanged into electrochemistry buffer using the PD10 column. UV-visible absorption spectroscopy was used to quantify protein concentration, and to determine the extent of protein oxidation during the course of sample preparation. Following this procedure, the protein was flash frozen in liquid nitrogen, and when necessary was transported under liquid nitrogen.
Flavodoxin II Kinetics on DDAB-Modified BPGE. The kinetics of the redox process observed on DDAB-modified BPGE were studied to determine if flavodoxin II diffused in the DDAB films on the time scale of voltammogram acquisition. Only the robust, lower potential species was monitored with this method. The signal intensity of the lower potential redox process observed in these films was proportional to the square root of the scan rate ( Figure S3C). Thus, the redox process occurring at the electrode was diffusion-limited, suggesting that flavodoxin II was   NaCl, and a cyclic voltammogram was acquired at a scan rate of 20 mV/s. The higher potential peak on the voltammogram had a midpoint potential of -138 ± 3. The lower potential peak on the voltammogram had a midpoint potential of -303 ± 1. The scan rate dependence of the signal intensity for free FMN in solution indicated that this molecule diffused in the DDAB films at a diffusion rate of 5.7 x 10 -7 ± 1.4 x 10 -7 cm 2 /s. This calculation assumed that the concentration of FMN was the same in solution as it was in the surfactant film, and that the redox process observed was a two electron process.

Figure S3. Cyclic Voltammogram of Flavodoxin II with the Higher and Lower
Potential Species. (S3A) The electrode was soaked in 200 µM authentic A. vinelandii flavodoxin II in 50 mM potassium phosphate, pH 7.5, 150 mM NaCl for 20 minutes. The electrode was transferred to 10 mM phosphate buffer, pH 7, and a cyclic voltammogram was acquired at a scan rate of 20 mV/s. Using this method, the midpoint potential of E 1 was measured as -454 ± 7 mV, and the midpoint potential of E 2 was measured as -164