Emulgen 913 was purchased from Kao Atlas (Osaka, Japan); all other chemicals were from Reakhim (Moscow, Russia). Ultrapure water was obtained using the Milli-Q system (Millipore, Bedford, USA).
Protein Expression and Purification
Bacteria were grown as previously reported  with slight modifications. Briefly, we used freshly transformed E. coli BL21DE3 to inoculate a preculture. The bacteria were allowed to grow in ampicillin-containing nutrient broth medium at 37°C overnight. These cultures were used to inoculate 4 l of a main culture containing ampicillin. Isopropyl-1-thio-D-galactopyranoside was added to induce heterologous protein production, and afterward cultures were grown at 37°C for 16 h. Recombinant Ad was purified after sonification as described, and the final concentration of Ad was determined using ε 414 = 9.8 mM-1 cm-1 . The purity of the Ad preparation was estimated by determining the relative absorbance of the protein at 414 and 273 nm, i.e. its Q value (A414/A273). AdR was heterologously expressed and purified as described elsewhere . The molar extinction coefficient used for estimation of AdR concentration was ε 450 = 10.9 mM-1 cm-1 . Isolation of CYP11A1 from bovine adrenal glands was performed as previously described . CYP11A1 concentration was estimated by carbon monoxide difference spectra using ε (450–490) = 91 (mM cm)-1.
Procedure for CYP11A1 Monomerization
For monomerization of cytochrome CYP11A1, the detergent Emulgen 913 in the concentration range 4–12% was chosen. The monomerization scheme was as follows: to 2 μl of stock solution of CYP11A1 (100 μM) in 50 mM KP, pH 7.4, were added 1.3 μl of Emulgen 913 at three various concentrations (10%, or 20%, or 30% solution) at T = 22°C. The final concentrations of Emulgen 913 in the three incubation solutions were 4, 8 and 12%, respectively. The mixture obtained was incubated at room temperature (22°C) for 10 min.
AFM Experiments and Samples' Preparation
AFM experiments were carried out using the direct surface adsorption method . As support, the mica was used.
For visualization of individual non-monomerized and monomerized CYP11A1 protein molecules, the appropriate protein solution was diluted in 50 mM K-phosphate buffer, pH 7.4 (50 KP) to obtain 1 μM protein concentration; 5 μl of obtained solution were immediately deposited onto the freshly cleaved mica surface and left for 3 min. For visualization of the individual Ad and AdR protein molecules, 5 μl of 1.0 μM solution of an appropriate protein in 50 mM K-phosphate buffer, pH 7.4, were deposited onto the freshly cleaved mica surface and left for 3 min. After that, each sample was first rinsed with the same buffer, then with ultrapure distilled water and dried in airflow. The binary complexes were obtained by mixing 10 μl of 5 μM solutions of appropriate individual proteins in 50 KP, pH 7.4. Then, the mixture was incubated for 10 min, diluted 2.5 times in the same buffer, and a 5-μl portion of the mixture was immediately placed onto mica. The ternary complexes were obtained by mixing 10 μl of 7.5 μM solutions of appropriate individual proteins in 50 KP, pH 7.4. Then the mixture was incubated for 10 min, diluted 2.5 times in the same buffer, and a 5-μl portion of the mixture was immediately placed onto mica. As was shown in an earlier research , with relative humidity exceeding 45%, the mica surface is covered with a water layer. Therefore, in the present study all the measurements were carried out at room temperature and at 60–70% air humidity, the protein molecules under study remained hydrated throughout. The choice of protein concentration was dictated by inherent limitations of the AFM technique: at higher concentrations, the molecules under observation formed layers on the mica support, which excluded the identification of individual objects.
All AFM experiments were carried out in a tapping mode on a multimode "NTEGRA" atomic force microscope (NT-MDT, Moscow, Russia) in air. Cantilevers NSG 10 produced by "NT-MDT" (Russia) were used. The resonant frequency of the cantilevers was 190–325 kHz, and the force constant was about 5.5–22.5 N/m. The calibration of the microscope by height was carried out on a TGZ1 calibration grating (NT-MDT, Moscow, Russia) with the step height 22 ± 0.5 nm. The supersharp probes with the radius of curvature of about 1–3 nm were used for measuring of CYP11A1 monomers' volumes. As supersharp probes, NSG01_DLC microprobes (NT-MDT, Russia) with a typical resonant frequency of 115–190 kHz were used.
The total number of measured particles in each sample was not less than 600, and the number of measurements for each sample was no less than 16, i.e. there were 4 measurements in each of the four series.
Analysis of AFM Images
The density of protein distribution with height, ρ(h), was calculated as ρ(h) = (N
/N) × 100%, where N
is the number of imaged proteins with height h, and N is the total number of imaged proteins. The calculation was carried out using a step of 0.2 nm.
To calculate the deaggregation degree, the dependence of distribution density ρ(h
) of CYP11A1 images with height (h
) was constructed:
The dependence of this distribution was approximated using root-mean square method by the sum of two curves:
are the parameters of ρ(h
) distribution. The maximum of ρ
(h) was calculated from Eq. (2).
For the analysis of distribution with heights and volumes (ρ(h
)) of imaged CYP11A1, (ρ(h
)) was calculated as
is the number of imaged proteins with the height h, and the volume V.
Values of height maximums and distributions widths, represented in text, were calculated from Eq. 2.
Photon correlation spectroscopy (PCS) measurements were carried out by use of N5 Submicron Particle Size Analyzer (Beckman Coulter, Inc). The principle of registration is based on measuring the interference pattern of light scattered on particles in solution by use of photon correlation spectroscopy (PCS). Measurements were made at the light-scattering angle of 90°. Protein solution (the stock one or the one subjected to monomerization procedure) was diluted in 50 mM KP, pH = 7.4, and placed into the measuring cuvette of Analyzer. Protein concentration was so selected as to make the intensity of dissipated light at 90° not lower than the sensitivity threshold corresponding to 5 × 104 counts. CYP11A1 and AdR concentrations were 5 μM for each protein. For Ad, the concentration was 0.2 mM. The measurements were made up to the accumulation of the signal during 200 s.
The calibration of the correlometer was performed using the set of latexes with the diameters 40, 50, 150 and 500 nm and the cytochrome C (2.9 × 5.5 × 2.3 nm) with the known X-ray structure from PDB . In this size range, the measured sizes of latex corresponded to nominal with a root-mean square deviation of 10%.
Optical Biosensor Measurements
Formation of the complex between monomeric CYP11A1 and Ad was additionally assayed on a Biacore 3000 system, using the optical biosensor method as described before with slight modifications [30, 31].
Briefly, after activation of the CM5 chip with N-ethyl-N'-dimethylaminopropyl-carbodiimide (EDC) and N-hydroxysuccinimide (NHS), 75 μL of a 200 μM Ad solution was injected with a flow of 5 μl min-1 at 20°C. The immobilization procedure was completed by injecting 1 M ethanolamine hydrochloride in order to block the remaining ester groups. Approximately 400 RU (response units) Ad was immobilized on the dextran matrix. In order to match the experimental conditions employed for the AFM measurements, we used a 50 mM potassium phosphate buffer (pH 7.4) containing 1% Emulgen 913. Binding of monomeric or oligomeric CYP11A1 to immobilized Ad was analyzed by injecting CYP11A1 solutions with concentrations varying between 1 and 100 nM. Each concentration was injected at least three times. To visualize unspecific background interactions between the dextran matrix and CYP11A1, a reference cell (i.e. the cell without Ad) was created. Ten microliters of 1 mM NaOH was used as regeneration solution. K
values were determined using the software Biaeval 4.1. Averaged binding curves for the interaction between Ad and varying CYP11A1 concentrations were fitted simultaneously using the 1:1 Langmuir-binding model. K
values were determined from the fit with the lowest standard deviation.
Control of Functionality of Monomeric CYP11A1
These assays were aimed toward demonstrating the functionality of monomeric CYP11A1. For this purpose, we investigated the conversion of 7-dehydrocholesterol to 7-dehydropregnenolone cortisol  using monomeric CYP11A1. In vitro reconstitution assays were performed as described before  with slight modifications. Briefly, the reaction mixture (0.5 ml) consisted of either CYP11A1 (0.4 μM) that has been monomerized using Emulgen 913 as described earlier or oligomeric CYP11A1 (0.4 μM), AdR (0.5 μM), Ad (0 to 4 μM), 7-dehydrocholesterol (400 μM) and MgCl2 (1 mM) in 50 mM potassium phosphate buffer (pH 7.4) containing 0.05% (v/v) Tween 20.
Substrate conversion was started by the addition of NADPH up to the final concentration of 100 μM. In addition to this, glucose-6-phosphate (5 μM) and glucose-6-phosphate dehydrogenase (1 U) were added to the reaction mixture. After the reaction was completed, steroids were extracted with chloroform and then separated on a Jasco reversed-phase HPLC system of the LC900 series using a 3.9 × 150 mm Waters Nova-Pak C18 column at 40°C. The mobile phase used for the separation was a mixture of acetonitrile/2-propanol (30:1). Product quantification was performed by correlating the product peak integrals with the peak area of a known internal standard (5 nmol cortisol) that was added prior to the chloroform extraction. K
and Vmax values were determined by plotting the substrate conversion velocity versus Ad concentration and applying the Michaelis–Menten kinetics (hyperbolic fit) using the program SigmaPlot 2001. Each experiment was performed four times. The velocity of the Ad-dependent product formation was expressed in nmol product × min-1 × nmol CYP11A1-1.
Proteins were analyzed via SDS gel electrophoresis in order to detect major impurities in protein preparations. The results obtained from these measurements revealed no impurities in the purified protein samples of all three components of the CYP11A1 electron transfer chain (data not shown).
In order to check possible structural changes in the protein conformations of the monomerized and oligomeric proteins, UV/VIS and CD spectroscopy have been performed.
Absorption spectra in the UV/VIS region (250–700 nm) were recorded at room temperature on a double-beam spectrophotometer UV2101PC (Shimadzu; Kyoto, Japan). UV/VIS spectra of monomeric or oligomeric proteins revealed no significant changes (data not shown). UV/VIS spectra of CYP11A1 displayed a pronounced peak at 392 nm, indicating that the protein is in its high spin conformation. Carbon monoxide difference spectroscopy performed for CYP11A1 displayed a pronounced peak at 450 nm, whereas the peak at 420 nm (non-functional protein) was not observable.
CD spectra of oxidized monomeric and oligomeric CYP 11A1 were recorded on a Jasco 715 spectropolarimeter as described before . All protein samples were diluted in 10 mM KP (pH 7.4). Possible changes in the secondary structures of the proteins were investigated by recording CD spectra in the range of 195–260 nm. CD measurements in the 250–650 nm range were performed using 10 μM proteins as described recently . The results obtained from these measurements revealed no significant conformational changes (data not shown) between the monomeric and oligomeric protein species.