Purification and functional characterization of human 11b hydroxylase expressed in Escherichia coli Andy Zo ¨ llner 1 , Norio Kagawa 1,2 , Michael R. Waterman 2 , Yasuki Nonaka 3 , Koji Takio 4 , Yoshitsugu Shiro 4 , Frank Hannemann 1 and Rita Bernhardt 1 1 Department of Biochemistry, Saarland University, Saarbru ¨ cken, Germany 2 Department of Biochemistry, Vanderbilt University School of Medicine, Nashville, TN, USA 3 College of Nutrition, Koshien University, Takarazuka, Hyogo, Japan 4 Biometal Science Laboratory, Riken Spring-8 Center, Harima Institute, Hyogo, Japan The final steps in the synthesis of the major human glucocorticoid, cortisol, and the most important miner- alocorticoid in humans, aldosterone [1], are catalyzed by 95% identical mitochondrial cytochrome P450 iso- zymes, 11b-hydroxylase (CYP11B1; EC 1.14.15.4) and CYP11B2 [2]. Cortisol is synthesized from 11-deoxy- cortisol through a hydroxylation reaction at position 11b catalyzed by CYP11B1, whereas aldosterone is synthesized from 11-deoxycorticosterone via a series of reactions catalyzed by CYP11B2. CYP11B1 is expressed in the adrenal zona fasiculata and is regulated by adrenocorticotrophic hormone (ACTH). The human aldosterone synthase, CYP11B2, on the other hand, is expressed in the zona glomerulosa Keywords Biacore measurements; congenital adrenal hyperplasia; expression; human CYP11B1; stopped-flow experiments Correspondence R. Bernhardt, Department of Biochemistry, Saarland University, 66123 Saarbru ¨ cken. Germany Fax: +49 681 302 4739 Tel: +49 681 302 3005 E-mail: ritabern@mx.uni-saarland.de Website: http://bernhardt.biochem.uni-sb.de/ HP1.html (Received 24 August 2007, revised 3 December 2007, accepted 18 December 2007) doi:10.1111/j.1742-4658.2008.06253.x The human 11b-hydroxylase (hCYP11B1) is responsible for the conversion of 11-deoxycortisol into the major mammalian glucocorticoid, cortisol. The reduction equivalents needed for this reaction are provided via a short elec- tron transfer chain consisting of a [2Fe-2S] ferredoxin and a FAD-contain- ing reductase. On the biochemical and biophysical level, little is known about hCYP11B1 because it is very unstable for analyses performed in vitro. This instability is also the reason why it has not been possible to stably express it so far in Escherichia coli and subsequently purify it. In the present study, we report on the successful and reproducible purification of recombinant hCYP11B1 coexpressed with molecular chaperones GroES ⁄ GroEL in E. coli. The protein was highly purified to apparent homogene- ity, as observed by SDS ⁄ PAGE. Upon mass spectrometry, the mass-to- charge ratio (m ⁄ z) of the protein was estimated to be 55 761, which is consistent with the value 55 760.76 calculated for the form lacking the translational initiator Met. The functionality of hCYP11B1 was analyzed using different methods (substrate conversion assays, stopped-flow, Bia- core). The results clearly demonstrate that the enzyme is capable of hydroxylating its substrates at position 11-beta. Moreover, the determined NADPH coupling percentage for the hCYP11B1 catalyzed reactions using either 11-deoxycortisol or 11-deoxycorticosterone as substrates was approx- imately 75% in both cases. Biacore and stopped-flow measurements indi- cate that hCYP11B1 possesses more than one binding site for its redox partner adrenodoxin, possibly resulting in the formation of more than one productive complexes. In addition, we performed CD measurements to obtain information about the structure of hCYP11B1. Abbreviations ACTH, adrenocorticotrophic hormone; AdR, adrenodoxin reductase; Adx, adrenodoxin; bCYP11B1, bovine 11b hydroxylase; hCYP11B1, human 11b hydroxylase. FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS 799 and is regulated by angiotensin II and potassium, with ACTH having mostly a short-term effect on expression [3,4]. Interestingly, in bovine CYP11B1 (bCYP11B1), which is the most widely studied CYP11B1 so far due to its availability, both functions of the human CYP11B isoforms (cortisol and aldosterone formation) are performed by a single protein. CYP11B1 deficiency results in decreased cortisol- production leading subsequently to an elevated plasma ACTH level, and an accumulation of steroid precur- sors. Such an enzyme deficiency is known to be the cause in 5–8% [5,6] of patients suffering from congeni- tal adrenal hyperplasia, an autosomal recessive inher- ited inborn error in steroidogenesis that ranks among the most frequent inborn errors of metabolism. CYP11B1 deficiency leads to flooding of the androgen synthesis pathway by accumulation of steroid precur- sors, resulting in hyperandrogenism. In approximately two thirds of patients, hypertension can be diagnosed because of the accumulation of 11-deoxycorticosterone and its metabolites. Overproduction of androgens such as in classic CYP11B1 deficiency leads, for example, to severe virilization of external genitalia in newborn females as well as to bone age acceleration in both sexes [2,6,7]. On the other hand, hCYP11B1 can also be a target of drug action in the case of hydrocortisolism, which plays an important role in the metabolic syndrome. So far, the development of selective and specific inhibitors was only possible using recombinant yeast or V79 cells for inhibition studies [8]. The reduction equivalents needed for all CYP11B1 ⁄ B2 catalyzed reactions are provided via a short electron transfer chain consisting of a [2Fe-2S] ferredoxin, adrenodoxin (Adx) and a NADPH-depen- dent, FAD containing reductase, adrenodoxin reduc- tase (AdR; EC 1.18.1.2) [8,9]. This electron transfer chain is also responsible for providing electrons for the conversion of cholesterol to pregnenolone, the precur- sor molecule of all steroid hormones, which is pro- duced in a reaction catalyzed by CYP11A1 [10,11]. So far, little is known about the interaction between hCYP11B1 and its redox partner Adx. This is mainly due not only to the scarce availability of human adre- nals, but also to the instability of this protein, which has hindered its expression in Escherichia coli and its subsequent purification. Therefore, most of the studies performed to date have been carried out using bovine CYP11B1 in a detergent solubilized system or in lipo- somes [12,13]. However, the instability of the solubi- lized enzyme, mainly due to its hydrophobic nature, has hindered any detailed investigation [13]. Moreover, purification of the homologous bovine protein from adrenal glands is known to be difficult, time consum- ing and renders only small quantities of the purified protein (4–8 mg from 1.25 g of mitochondrial pellets [14]). In the present study, we describe the successful expression of human CYP11B1 in E. coli as well as its subsequent purification in significant quantities. Addi- tionally, we were able to functionally characterize this enzyme by using bovine Adx and AdR as electron donors. Taking this into account, this study opens new perspectives for the investigation of the structure and functions of this physiologically important protein. Results and Discussion Previously, rat CYP11B1 and CYP11B2 was expressed in E. coli JM109 using a bacterial expression vector pTrc99A [15]. However, the expression level of the proteins was 10–20 nmolÆL )1 culture media, which is too low to obtain quantities of the purified proteins for in depth characterization. As it was very efficacious for the expression of human CYP19 [16], mouse CYP27B1 [17] and bovine CYP21 [18], the coexpression of molecular chaperones GroES ⁄ GroEL also resulted in an efficient expression of human CYP11B1. Utilizing the pET⁄ BL21 expres- sion system with the coexpression of molecular chaper- ones GroES ⁄ GroEL, the human CYP11B1 has been expressed in E. coli with a yield of approxi- mately 400 nmolÆ L )1 culture. In addition to greatly increasing the expression level, our expression system using pET ⁄ BL21 shortened the incubation time to approximately 24 h compared to 45 h for the pTrc99A ⁄ JM109 system used for the expression of rat CYP11B1 and CYP11B2 [15]. The expressed form of hCYP11B1 was stable in the presence of detergents and glycerol and highly purified through three chromatographic steps to the specific content of 19.8 nmolÆmg )1 , as estimated from the reduced CO-difference spectrum and protein assay (the theoretical value 17.8 nmolÆmg )1 ). The purified protein was apparently homogeneous upon SDS ⁄ PAGE (Fig. 1) and showed a single major peak on HPLC analysis using a POROS column (Fig. 2A). The peak was collected and subjected to MALDI-TOF analysis. Signals of singly (m ⁄ z = 55761) and doubly (m ⁄ z = 27898) charged apoprotein were observed (Fig. 2B). The m ⁄ z value is in good agreement with the calculated molecular mass of 55760.76 for the transla- tional initiator Met-deleted hCYP11B1. As shown in Fig. 3, the UV ⁄ visible spectrum of purified recombinant CYP11B1 revealed a pronounced Soret peak at 392 nm in the absence of substrates (Fig. 3, spectrum 1), indicating that the protein is in its Functional characterization of hCYP11B1 A. Zo ¨ llner et al. 800 FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS high spin state. The spectrum was not changed by the addition of 17,21-hydroxyprogesterone (spectrum 2), although the P450 was reduced by sodium dithionite (spectrum 3) and formed the reduced CO complex (spectrum 4) that produced a typical P450 peak of reduced CO-difference spectrum at 448 nm (Fig. 3, lower panel). The finding that the recombinant protein is in its five coordinated high spin state [9,19] is of importance because it has been postulated that the high spin state of cytochrome P450s is more stable compared to its low spin state probably due to slight conformational differences of the active site. Additional spectroscopic measurements have been performed using CD spectropolarimetry. The CD spectrum of 11B1 in the far-UV region (Fig. 4A) describes a mainly alpha helical conformation, as expected for a cytochrome P450 enzyme because all P450 structures solved to date are predominantly alpha helical. The spectrum of CYP11B1 is characterized in this region by a negative dichroic double band with minima at 210 and 221 nm, which do not change after substrate supplementation (data not shown). The heli- cal content of the protein at 20 °C was determined to be greater than 50% according to the contin and selcon prediction programs [20,21]. In the near-UV and visible region, the CD spectra of CYP11B1 display two large signals of negative sign (Fig. 4B), one with a minimum below 280 nm and a second with a minimum near the position of the Soret maximum (386 nm) in the absorption spectrum. The negative cotton effect in the Soret region has been observed in other cyto- chromes P450 and can be attributed to a solvent-acces- sible heme pocket. In addition, two signals of positive sign appeared at 290 and 326 nm. The 260–280 nm region reflects mainly tyrosine transitions, whereas the signal at 326 can be attributed to an anisotropy of the porphyrin absorption band [22]. Upon addition of substrate, the signal of the positive CD band at 290 nm and the negative band at 386 nm decreased (Fig. 4B). A similar observation was described for sub- strate binding of cytochrome P450RR1 from Rhodo- coccus rhodochrous [23]. The functionality of the enzyme was demonstrated by performing hCYP11B1-dependent substrate conver- sion assays using 11-deoxycorticosterone and 11-de- oxycortisol as substrate and by a subsequent HPLC analysis of the steroid product pattern (Fig. 5). The hCYP11B1 electron transfer chain was always reconsti- tuted using bovine AdR and bovine Adx, which is Purified human CYP11B1 24 µg 8 Fig. 1. SDS polyacrylamide gel electrophoresis of the purified CYP11B1. Different amounts of the purified hCYP11B1 (2, 4, 8 lg per lane) were separated by SDS ⁄ PAGE (10%) and visualized by Coomassie staining. A 0 5 10 15 20 25 20 40 27898 55761 60 80 100 Time (min) m/z (×10 3 ) B Fig. 2. HPLC and mass spectral analysis of the purified hCYP11B1. (A) The purified hCYP11B1 was applied on RP-HPLC analysis using POROS R1 ⁄ 10 (2.1 · 100 mm; Applied Biosystems) as described in Experimental procedures. The protein absorbance was monitored at 215 nm. (B) The apoprotein peak (15.5 min peak in Fig. 2A) was collected and subjected to MALDI-TOF MS with sinapic acid as matrix. A. Zo ¨ llner et al. Functional characterization of hCYP11B1 FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS 801 capable of interacting with hCYP11B1 possessing a 90% amino acid sequence identity with human Adx. As shown in Fig. 6, the recombinant enzyme was able to efficiently convert 11-deoxycortisol to cortisol with a k cat of 1.67 s )1 and a K m of 338.4 ± 30.2 lm for 11-deoxycortisol. Human CYP11B1 is also able to convert 11-deoxycorticosterone to corticosterone (k cat = 0.85 s )1 and K m = 179.5 ± 19.1 lm 11-deoxy- corticosterone). The fact that the binding affinity of hCYP11B1 to 11-deoxycorticosterone is significantly higher compared to the affinity for its natural substrate, 11-deoxycorti- sol, might be caused by the additional hydroxyl group at position C17 of 11-deoxycortisol. This additional OH group is likely to hinder the entrance of the slightly more hydrophilic and bulky substrate 11-de- oxycortisol into the active site of the enzyme. Compared with the published values for the forma- tion of corticosterone by the bovine enzyme, bCYP11B1 (k cat = 0.1 s )1 ), the obtained values using hCYP11B1 are approximately ten-fold higher. This finding is quite surprising because the sequence identity between the human enzyme and the bovine enzyme is high (73%). However, some of the main differences between human and bovine CYP11B1 are located in 0.15 0.10 0.05 AbsorbanceΔAbsorbance 1 2 3 4 1: Free 2: 17,21-OH-Prog 3: Reduced 4: Reduced-CO 0.00 0.15 0.10 0.05 0.00 -0.05 350 400 450 500 4-3 550 600 650 700 350 400 450 500 550 Wavelength (nm) 600 650 700 Fig. 3. UV ⁄ visible spectra of the purified human CYP11B1. The absolute spectra of the purified CYP11B1 (0.15 l M) was analyzed without substrate (1), with 0.1 m M of 17,21-hydroxy-progester- one (2), as a reduced form with 17,21-hydroxy-progesterone in the presence of sodium dithionite (3), and as a reduced CO complex with 17,21-hydroxy-progesterone (4). The reduced CO-different spectrum (4-3) in the presence of 17,21-hydroxy-progesterone is shown in the lower panel. B A Fig. 4. CD spectra of CYP11B1. The CD spectrum recorded in the far-UV of CYP11B1 is shown in the absence of substrate at 20 °C (A). The protein concentration was 5 l M in 10 mM potas- sium phosphate buffer, pH 7.4. CD spectra in the near ultraviolet and visible light were recorded in the absence (black) and in the presence of substrate 11-deoxycortisol (gray) at 20 °C (B). The con- centration of CYP11B1 was 10 l M in 10 mM potassium phosphate buffer, pH 7.4. Substrate was added to a concentration of 20 l M. Functional characterization of hCYP11B1 A. Zo ¨ llner et al. 802 FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS substrate recognition sites 2 and 3 (SRS2 and SRS3), which are mainly composed of the F and G helix of the cytochrome [24]. This inconsistency might be the cause for the significantly different interspecies conver- sion rates. However, the alterations may also be explained by the broader substrate-binding spectrum of bCYP11B1 resulting from the combination of the functions of CYP11B1 and CYP11B2 within one enzyme. Comparison of the k cat values obtained for the con- version of 11-deoxycorticosterone by hCYP11B1 with values published for corticosterone formation by rec- ombinantly expressed and purified CYP11B1 from rats (2.18 s )1 ) indicated that the maximal rate determined for hCYP11B1 has not been significantly altered (approximately 2.5-fold lower). The k cat values obtained for hCYP11B1 catalyzed reactions are in the range of values previously reported for CYP106A2 catalyzed steroid hydroxylations and significantly faster than the k cat value determined for the formation of pregnenolone from cholesterol cata- lyzed by CYP11A1, which is the rate limiting step in steroidogenesis. Additional studies performed to correlate the NADPH consumption during the reaction with the amount of product formed (i.e. the coupling percent- age) revealed a 72% coupling when using 11-deoxy- corticosterone as substrate and a 76% coupling when using 11-deoxycorticosterone as substrate, respectively. These findings indicate that, during the reaction of hCYP11B1 with either 11-deoxycorticosterone or 11-deoxycortisol, approximately 25% of the consumed NADPH is spent in other reactions (e.g. hydrogen BA Fig. 5. Typical steroid product pattern recorded at 440 nm after HPLC separation with an isocratic solvent system consisting of acetonitrile ⁄ water (60 : 40) at a flow rate of 1 mLÆmin )1 on a C18 reversed phase col- umn. Substrate conversions of CYP11B1 were performed with increasing amounts of Adx (gray, dark gray, black lines). (A) Conversions of the substrate 11-deoxy- corticosterone (DOC) to the products B (cor- ticosterone) and 18OH-B using cortisol (F) as internal standard. (B) shows the conver- sion of the substrate 11-deoxycortisol (RSS) to F using 11-deoxycorticosterone as inter- nal standard. A B k k m m Fig. 6. hCYB11B1 substrate conversion assays were performed using the reconstituted electron transfer chain consisting of bovine Adx and bovine AdR as well as different substrate concentrations: 11-deoxycortisol (A) and 11-deoxycorticosterone (B). Steroid separa- tion was achieved via HPLC analysis as indicated in Experimental procedures. V max values (nmol productÆmin )1 Ænmol )1 hCYP11B1) were subsequently converted into k cat values (s )1 ). A. Zo ¨ llner et al. Functional characterization of hCYP11B1 FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS 803 peroxide formation). Besides the formation of hydro- gen peroxide, other NADPH consuming reactions probably involving the redox partners AdR and Adx are taking place. Further studies will focus on investi- gating this interesting question. Experiments performed to determine the Adx depen- dency of the hCYP11B1 catalyzed reaction were car- ried out under substrate saturation conditions. The k cat values obtained in these experimental set ups were in the range of the values shown above using different substrate concentrations (Table 1). The Adx-dependent K m values obtained from these studies using 11-deoxy- corticosterone or 11-deoxycortisol as substrate did not reveal any significant differences, indicating that the interaction between Adx and hCYP11B1 is not affected by the different substrates (Table 1). In addi- tion to this, the obtained Adx-dependent K m value using 11-deoxycorticosterone as substrate was in the range of values determined in previous studies for the interaction between bovine Adx and bCYP11B1. Biacore measurements were performed to investigate the binding behavior between bovine Adx ox and hCYP11B1 ox or bCYP11B1 ox in more detail. Among the binding models available in the standard software (e.g. 1 : 1 binding or complexes with higher stoichio- metry), the best fit was always observed with the ‘bivalent analyte’ model. This suggested that CYP11B1 possesses more than one binding site for Adx. Taking possible steric hindrances on the chip surface into account, only the formation of the first predominant 1 : 1 complex has been considered (Table 2), as was the case in a previous study [25]. As seen in Table 2, the K D values obtained for the predominant 1 : 1 com- plexes for both CYP11B1 species were in the nm range. Surprisingly, the k on rate of the bCYP11B1 ⁄ Adx com- plex was two-fold slower compared with the on-rate for the hCYP11B1 ⁄ Adx complex. On the other hand, the off rate was five-fold faster for the hCYP11- B1 ⁄ Adx, indicating a slightly weaker stability com- pared to the physiological interaction. To characterize the electron transfer from Adx to hCYP11B1, we performed stopped flow experiments. The recorded reaction traces displayed three phases that could be fitted separately (Fig. 7). This might indi- cate that there are three different Adx binding sites on hCYP11B1 or that complex rearrangements leading to different reduction rates take place during the reaction. The k obs values determined for the first phase were always in the range of 60 s )1 , indicating a fast process. However, the amplitude of this phase was only approximately 15–20% of the overall reaction ampli- tude, indicating that this productive complex might be thermodynamically less favored. Due to the velocity and the small amplitude change, it was not possible to determine the Adx dependency of this first process. Both the second phase and the third phase displayed an Adx dependency of the k obs rate, which could be evaluated using the Michaelis–Menten equation (Fig. 7). Combined with the data obtained from Bia- core experiments, which indicated the possibility of more than one complex formation, and considering that no impurities could be detected in polyacrylamid gel electrophoresis using our hCYP11B1 preparation, it is unlikely that the observed phases are a result of a heterogenous sample composition. Since it is known that the Adx concentration plays a role in the regula- tion of the activity of CYP11A1 [26], CYP11B1 [27] and CYP11B2 [28], this finding is not surprising. The maximal velocities extracted from the plots shown in Fig. 7B,C for the second and third phase were 3.89 s )1 and 0.65 s )1 , respectively. The K D values obtained from these experiments for the interaction between the relevant redox states of Adx and hCYP11B1 were 0.78 lm for the second phase and 2.2 lm for the third phase. Since the K D value obtained from the optical Table 1. Kinetic parameters obtained for the conversion of 11-deoxycorticosterone or 11-deoxycortisol using different Adx concentrations under substrate saturation (left). Kinetic parameters obtained using different substrate concentrations are also shown (right). Substrate Adx-dependent Substrate-dependent K m (lM) k cat (s )1 ) K m (lM) k cat (s )1 ) 11-Deoxycorticosterone 2.0 ± 0.21 1 ± 0.06 180 ± 19 0.85 ± 0.07 11-Deoxycortisol 2.4 ± 0.5 1.48 ± 0.08 338 ± 30 1.67 ± 0.14 Table 2. Values obtained for the complex formation between bovine Adx and CYP11B1 from different species using a Biacore 3000 system. Values were determined using a bivalent mechanism. The values shown below characterize the formation of the predomi- nant 1 : 1 complex. The 2 : 1 complex was not considered (see Results and Discussion). Complex k on (s )1 ÆM )1 ) k off (s )1 ) K D (M · 10 )6 ) bAdx- ⁄ hCYP11B1 775 000 0.1089 0.141 bAdx ⁄ bCYP11B1 300 000 0.019 0.063 Functional characterization of hCYP11B1 A. Zo ¨ llner et al. 804 FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS biosensor measurements and the value obtained for the productive complex leading to the second reaction phase are in the same range, it can be assumed that this complex is the predominant complex seen with the Biacore device. Moreover, analysis of the amplitude change of these phases extracted from the stopped-flow experiments indicates that the second complex is fur- ther stabilized in the presence of increasing amounts of Adx, whereas the complex leading to the third reaction phase is favored in the presence of small amounts of Adx (Fig. 7). These findings suggest different thermo- dynamic attributes for the productive complexes. Nevertheless, it cannot be ruled out that the differ- ent reaction phases observed in these experiments are caused by complex rearrangements or conformational gating that might be necessary before an efficient elec- tron transfer can take place. More investigations including Adx and CYP11B1 mutants will be necessary to investigate this question in more detail. However, considering the current data, it is very likely that a productive interaction (complex formation) between Adx and CYP11B1 can take place through more than one productive complex, as previously postulated for the interaction between bovine CYP11B1 and Adx [29]. Moreover, comparison of the k cat values obtained from the substrate conversion experiments with the maximal observed reduction rates from the stopped- AB CD Fig. 7. Stopped flow analysis of the Adx-dependent hCYP11B1-CO complex formation. The transient reaction traces obtained for this inter- action displayed three different phases that could be evaluated by using mono- or biexponential fits. (A) k obs values for the first fast phase using different Adx concentrations were obtained after evaluation using a monoexponential function as shown in the insert. (B) k obs values obtained for the second phase plotted against the corresponding Adx concentration. k obs,max and K m values were determined by using a hyperbolic equation as shown in the plot. The insert shows the curve trajectory excluding the first phase, which was best described by a bi- exponential fit. (C) Plot showing the Adx dependency of the kobs values obtained for the third reaction phase along with the deter- mined k obs,max and K m values (D) Adx concentration-dependent amplitude change of the different reaction phases expressed as a percent. A. Zo ¨ llner et al. Functional characterization of hCYP11B1 FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS 805 flow experiments indicates that only the first fast phase can enable such high turnover rates. Additionally, it appears that the second and the third phase seen in the stopped-flow measurements are negligible during the hydroxylation reaction, although possessing a higher amplitude change in the stopped-flow measure- ments compared to phase 1. Otherwise, the k cat values from the substrate conversion assays were likely to be in the range of the k obs,max values obtained for these phases. However, more investigations are necessary to clarify this assumption. In this context, it is possible that the predominant, but slower phases observed in the stopped-flow experiments play a role in the regula- tion of the activity of CYP11B1, especially since the absolute amplitude change of these phases when using higher Adx concentrations increases and an involve- ment of the Adx concentration in the regulation of CYP11B1 has been demonstrated previously [27]. Nev- ertheless, the physiological relevance of these different phases and the postulated different complexes remains unclear and should be subject of further studies. In summary, the present study reports on the suc- cessful purification of functional hCYP11B1 expressed in E. coli. This will open new possibilities for analyzing this very important cytochrome P450 in vitro, including the detailed investigation of the interaction of hCYP11B1 with its redox partner, Adx. As indicated by stopped-flow and optical biosensor experiments, it is very likely that the reaction between hCYP11B1 and Adx can proceed through more than one productive complex. In addition, the purification protocol provided here will facilitate the examination of mutations in hCYP11B1 that lead to congenital adrenal hyperplasia, indicating the medical relevance of the present study. To date, the examination of such hCYP11B1 mutants has been only possible through cell culture experi- ments, which do not provide detailed information on the influence of such mutations on the protein struc- ture or on its redox behavior. Finally, the expression and purification of hCYP11B1 is a necessary prerequi- site for its future structural characterization. Experimental procedures Protein expression and purification The human CYP11B1 was expressed as a mature form with N- and C-terminal modifications. The cDNA fragment encoding the modified CYP11B1 was produced by PCR using the 5¢-primer (CGCCATATGGCTACTAAAGCTG CTCGTGTTCCACGTACAGTGCTGCCA) and 3¢-primer (GCGAAGCTTAATGATGATGATGATGATGGTTGAT GGCTCTGAAGGTGAGGAG) and inserted into the NdeI ⁄ HindIII-digested pET17b expression vector. The cDNA template was as described previously [30]. The DNA sequence was determined by automated sequencing. This 5¢-primer was designed to alter the N-terminus from the original GTRAAR– – – to MATKAAR– – –. The 3¢-primer was designed to add six histidine residues at the C-terminus to facilitate the purification. The CYP11B1 expression plasmid was introduced into E. coli strain BL21(DE3)pLys along with a GroES ⁄ GroEL expression vector pGro12 [31]. The human CYP11B1 was expressed and extracted from E. coli in a similar manner to the methods previously described for the expression of CYP19 and CYP21 [18]. The extracts (100 mL) from 1 L culture were applied on a Ni-NTA agarose (10 mL bed volume) column equilibrated with buffer A (50 mm potassium phosphate, pH 7.4, 500 mm sodium acetate, 20% glycerol, 0.1 mm EDTA, 0.1 mm dithiothreitol, 1% sodium cholate, 1% Tween 20, 0.1 mm phenylmethanesulfonyl fluoride), washed with 75 mL of buffer A plus 40 mm imidazole, and with 20 mL of buffer B (50 mm potassium phosphate, pH 7.4, 20% glycerol, 0.1 EDTA, 0.1 mm dithiothreitol, 40 mm imidazole, 1% sodium cholate, 1% Tween 20, 0.1 mm ATP and 0.1 mm phenylmethanesulfonyl fluoride). Proteins were eluted with buffer C (200 mm imidazole acetate, pH 7.4, 20% glycerol, 0.1 mm EDTA, 0.1 mm di- thiothreitol, 1% sodium cholate, 1% Tween 20). The red- colored fractions were combined and diluted with five volumes of buffer D (20% glycerol, 0.1 mm EDTA, 0.1 mm dithiothreitol, 1% sodium cholate, pH 7.4) and applied on a DEAE-Sepharose (30 · 50 mm) equilibrated with buffer E (20 mm potassium phosphate, pH 7.4, 20% glycerol, 0.1 mm EDTA, 0.1 m m dithiothreitol, 10 mm imidazole, 1% sodium cholate, 0.1% Tween 20). The column was washed with 40 mL of buffer E. Pass- through fractions were then applied on a SP-Sepharose column (30 · 40 mm) equilibrated with buffer F (20 mm potassium phosphate, pH 7.4, 20% glycerol, 0.1 mm EDTA, 0.1 mm dithiothreitol, 10 mm imidazole, 1% sodium cholate). The column was washed with 20 mL of buffer F, and eluted with 0–125 mm NaCl gradi- ent in buffer G (40 mm potassium phosphate, pH 7.4, 20% glycerol, 0.1 mm EDTA, 0.1 m m dithiothreitol, 10 mm imidazole, 1% sodium cholate). The major red frac- tions were combined, and repeatedly concentrated and diluted using a centrifugal device to replace the buffer with buffer H (50 mm potassium phosphate, pH 7.4, 20% glyc- erol, 0.1 mm EDTA, 0.1 mm dithiothreitol, 1% sodium cholate, 0.05% Tween 20). The protease deficient E. coli strain BL21 was used as host strain for the heterologous expression of AdR and Adx. The plasmid containing the coding sequence for AdR was kindly provided by Y. Sagara [32]. Recombinant Adx and AdR were purified as described previously [33,34]. Functional characterization of hCYP11B1 A. Zo ¨ llner et al. 806 FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS Isolation of CYP11B1 from bovine adrenals was performed as described by Ikushiro et al. [13] with slight modifica- tions. RP-HPLC and mass spectrometry of the purified hCYP11B1 Purity of the h11B1 was assessed by RP-HPLC. RP-HPLC was conducted on a column Poros R1 ⁄ 10 (2.1 · 100 mm; Applied Biosystems, Foster City, CA, USA) using a liquid chromatograph (Agilent model 1100; Agilent Technologies, Palo Alto, CA, USA) with a 16 min linear gradient of 8–72% CH3CN in 0.1% trifluoroacetic acid at a flow rate of 0.1 mLÆmin )1 . Column effluent was monitored by absor- bance at 215 nm, 254 nm, 275 nm, 290 nm and 400 nm. The peak eluted at 13.5 min contained the heme extracted from h11B1 and that at 15.5 min contained the apoprotein. The apoprotein peak was collected and subjected to MALDI-TOF MS to verify the integrity of the protein moiety on Voyager DE-Pro (Applied Biosystems) with sina- pic acid as matrix. UV/visible and CD spectroscopy UV ⁄ visible spectra of CYP11B1 were recorded at room temperature on a Shimadzu double-beam spectrophoto- meter (UV2100PC; Shimadzu, Kyoto, Japan). The con- centration of the 11b-hydroxylase was estimated by carbon monoxide difference spectra assuming e 450–490 = 91 mm )1 Æcm )1 according to [35]. Adx and AdR concentra- tions were determined using the molar extinction coefficient e 415 = 9.8 mm )1 Æcm )1 [36] and e 450 = 10.9 mm )1 Æcm )1 [37], respectively. CD spectra were recorded at 20 °C on a Jasco J720 spec- tropolarimeter (Jasco Corporation, Tokyo, Japan). Samples contained 10 lm CYP11B1 in 10 mm potassium phosphate buffer (pH 7.4) in a 1 cm cuvette for measurements in the 250–650 nm range and 5 lm CYP11B1 in the same buffer in a 0.1 cm cuvette for measurements in the 190– 250 nm range. The spectra were accumulated five times and then smoothed. The spectrum of the potassium phosphate buffer was recorded in each case as a baseline. Substrate 11-deoxycortisol was added to a concentration of 20 lm. Secondary structure content analysis was performed using the contin and selcon programs [20,21]. Enzyme activity assays These assays served the purpose to demonstrate the ability of the recombinant CYP11B1 enzyme to 11b-hydroxylate its natural substrates, 11-deoxycortisol and 11-deoxycorti- costerone, to form cortisol and corticosterone, respectively. Assays aimed at the characterization of the CYP11B1 activ- ity depending on the Adx concentration were performed as previously described for CYP11A1 reconstitution assays [38] with slight modifications. All experiments were per- formed using bovine Adx, which is capable of interacting with hCYP11B1. Bovine and human Adx exhibit a 90% primary structure identity. Briefly, the reaction mix- ture (0.5 mL) consisted of CYP11B1 (0.4 lm), AdR (0.5 lm), Adx (2–20 lm), 11-deoxycortisol or 11-deoxycorti- costerone (400 lm), MgCl 2 (1 mm)in50mm Hepes buffer (pH 7.3, 0.05% (v ⁄ v) Tween 20). In addition to this, a NADPH regenerating system consisting of MgCl 2 (1 lm), glucose 6-phosphate (5 lm) and glucose 6-phosphate dehy- drogenase (1 U) was applied. In another set of experiments, we varied the substrate concentration in the range 0–700 lm for both substrates whereas the Adx concentration was fixed at 10 lm. All other components were as described above. All reactions were ini- tiated by the addition of 100 lm NADPH and were carried out for 10 min at 37 °C. After stopping the reaction by add- ing chloroform, either cortisol (for 11-deoxycorticosterone conversion assays) or 11-deoxycorticosterone (for 11-deoxy- cortisol conversion) was added to the corresponding reac- tion mixture as an internal standard. After extraction of the steroids and evaporation of the chloroform phase, the ste- roids were resuspended in 200 lL acetonitrile and separated on a Jasco reversed phase HPLC system of the LC900 series using a 3.9 · 150 mm Waters Nova-Pak C 18 column (Waters Corporation, Milford, MA, USA). Column temper- ature was kept constant at 25 °C with a peltier oven. The mobile phase used for steroid separation was a mixture of acetonitrile ⁄ water (60 : 40) at a flow rate of 1 mLÆmin )1 . Steroid separation was monitored at 240 nm over a period of 3 min. Product quantification was performed by correlat- ing the product peak area with the peak area of the known internal standard steroid (5 nmol cortisol or 11-deoxycorti- costerone) added prior to the chloroform extraction. K m and V max values were determined by plotting the sub- strate conversion velocities versus the corresponding Adx or substrate concentrations and by subsequently applying Michaelis–Menten kinetics (hyperbolic fit) using the pro- gram sigmaplot 2001 (Systat Software, San Jose, CA, USA). Extracted V max values (nmol productÆmin )1 Æ nmol )1 hCYP11B1) were subsequently converted into k cat values (s )1 ). To correlate the NADPH consumption with the amount of product formed (i.e. the coupling percentage), we per- formed additional experiments. Samples generated for this purpose contained 400 l m substrate (11-deoxycortisol or 11-deoxycorticosterone), 0.4 lm hCYP11B1, 3 lm Adx, 0.5 lm AdR in 50 mm Hepes buffer, pH. 7.4, containing 0.05% Tween 20. The reaction was initiated by addition of NADPH to a final concentration of 100 lm. Reaction con- ditions were as described above. The sample volume was 500 lL. NADPH consumption was determined spectro- scopically by recording the absorption changes of the A. Zo ¨ llner et al. Functional characterization of hCYP11B1 FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS 807 reaction mixture at 340 nm, corresponding to the absorp- tion maximum of NADPH, at the start of the reaction (t = 0) and after 10 min. To subtract protein absorption at this wavelength, we used a reference reaction sample with- out NADPH. NADPH consumption (i.e. the amount of NADPH consumed during the reaction) was sub- sequently determined by using the Lambert–Beer law (e 340 NADPH = 6.3 mm )1 Æcm )1 ) and the sample volume. Product formation was determined as described above by stopping the reaction with chloroform after 10 min and subsequently separating the extracted steroids via HPLC. Product quantification was performed as described before. NADPH consumption values were subsequently correlated with the amount of product formed to provide coupling values expressed in percent (i.e. amount of product in nmol · consumed amount of NADPH in nmol )1 · 100). Optical biosensor measurements Formation of the bAdx ⁄ bCYP11B1 and bAdx ⁄ hCYP11B1 complexes was assayed on a Biacore 3000 system (Biacore, Uppsala, Sweden), using the optical biosensor method described previously [39] with slight modifications. After activation of the CM5 chip with N-ethyl-N¢-di- methylaminopropyl-carbodiimide and N-hydroxysuccini- mide, 75 lL of a 200 lm Adx solution was injected with a flow of 5 lLÆ min )1 at 15 °C. The immobilization procedure was completed by injecting 1 m ethanolamine hydrochloride to block the remaining ester groups. Approximately 400 response units of Adx were immobilized on the dextran matrix. Binding of hCYP11B1 or bCYP11B1 was analyzed after injection of solutions with varying concentrations in the range 10–500 nm. Each concentration was injected at least three times. To visualize unspecific background inter- actions between the dextran matrix and CYP11B1, a refer- ence cell was created. To remove the bound CYP11B1, 10 lLof a 2mm NaOH solution was injected. K D values were determined using the software biaeval, version 3.1. Averaged binding curves for the interaction between Adx and varying CYP11B1 concentrations were fitted simulta- neously using different binding models available in the eval- uation software (e.g. 1 : 1 Languimir-binding or a bivalent binding model as at least two possible interaction sites for Adx exist on CYP11B1). K D values were determined from the fit with the lowest standard deviation. Kinetics by rapid mixing Stopped flow measurements were carried out on a SFM 300 stopped-flow spectrophotometer equipped with a FC 100 ⁄ 10 cuvette and a MPS 60 data-processing unit (Biologic SAS, Claix, France) at 15 °C. Anaerobic conditions were achieved by incubation of the stopped-flow device for 20 min with argon-bubbled buffer containing 5 mm dithionite followed by repeated flushing with excessively Ar-bubbled reaction buffer to remove oxygen and remaining dithionite from the system. All samples were prepared in a glove box in an oxy- gen-free atmosphere. The reaction buffer applied for all mea- surements was a 50 mm Hepes buffer (pH 7.4) containing 0.05% Tween 20 [39]. To follow the reduction of cytochrome CYP11B1 by AdR-reduced Adx, the absorption changes were monitored at 450 nm, which corresponds to the formation of the ferrous–carbon monoxide complex [40–43] as previously described for measurements carried out with CYP11A1 [29,39]. Prior to mixing, syringe A contained CYP11B1 (2 lm) whereas syringe B was filled with NADPH (200 lm), AdR (2 lm) and varying concentrations of Adx in the range 0.5–32 lm. The mixture in syringe B was allowed to age for 5 min to assure complete reduction of Adx. The solutions in the two syringes were saturated with CO prior to loading into the driving syringes. All resulting curves were evaluated using sigmaplot 2001. Kinetic traces were analyzed using monoexponential or biexponential fits to extract corresponding reduction rates. The k obs values were plotted against the corresponding Adx concentration and the curve was fitted with a hyperbolic equation to extract maximal reduction rates, k obs,max . These plots were also used to obtain K D values for the interaction between the relevant redox states of the reacting proteins. Acknowledgements This work was supported by a grant from the Fonds der Chemischen Industrie to RB and GM37942 to MRW. The authors would like to thank K. Neumann, A. Eiden-Plach and W. Reinle for their excellent tech- nical support. References 1 Miller WL & Tyrell JB (1995) The adrenal cortex. In Endocrinology and Metabolism (Felig P, Baxter J & Frohman L, eds), pp. 555–711. McGraw-Hill Press, New York, NY. 2 White PC, Curnow KM & Pascoe L (1994) Disorders of steroid 11-beta-hydroxylase isozymes. Endocr Rev 15, 421–438. 3 Mornet E, Dupont J, Vitek A & White PC (1989) Char- acterization of two genes encoding human steroid 11 beta-hydroxylase (P-450(11) beta). J Biol Chem 264 , 20961–20967. 4 Kawamoto T, Mitsuuchi Y, Toda K, Miyahara K, Yo- koyama Y, Nakao K, Hosoda K, Yamamoto Y, Imura H & Shizuta Y (1990) Cloning of cDNA and genomic DNA for human cytochrome P-45011 beta. FEBS Lett 269, 345–349. 5 Zachmann M, Tassinari D & Prader A (1983) Clinical and biochemical variability of congenital adrenal-hyper- Functional characterization of hCYP11B1 A. Zo ¨ llner et al. 808 FEBS Journal 275 (2008) 799–810 ª 2008 The Authors Journal compilation ª 2008 FEBS [...]... to 11-beta -hydroxylase deficiency – a study of 25 patients J Clin Endocrin Metab 56, 222–229 Krone N, Grischuk Y, Muller M, Volk RE, Grotzinger J, Holterhus PM, Sippell WG & Riepe FG (2006) Analyzing the functional and structural consequences of two point mutations (P94L and A368D) in the CYP11B1 gene causing congenital adrenal hyperplasia resulting from 11 -hydroxylase deficiency J Clin Endocrin Metab... Direct expression of adrenodoxin reductase in Escherichia coli and the functional characterization Biol Pharm Bull 16, 627–630 Uhlmann H, Beckert V, Schwarz D & Bernhardt R (1992) Expression of bovine adrenodoxin in E coli and site-directed mutagenesis of ⁄ 2 Fe-2S ⁄ cluster ligands Biochem Biophys Res Commun 188, 1131–1138 Omura T & Sato R (1964) The carbon monoxide-binding pigment of liver microsomes... M, Sandhoff K & Hasilik A (2004) Purified recombinant human prosaposin forms oligomers that bind procathepsin D and affect its autoactivation Biochem J 383, 507–515 Bernhardt R (2000) The role of adrenodoxin in adrenal steroidogenesis Curr Opinion Endocrinol Diabetes 7, 109–115 Hampf M, Dao NT, Hoan NT & Bernhardt R (2000) New molecular genetic defects causing 11betahydroxylase deficiency (CAH) Endocr... Characterization of stable human aromatase expressed in E coli Steroids 69, 235–243 Uchida E, Kagawa N, Sakaki T, Urushino N, Sawada N, Kamakura M, Ohta M, Kato S & Inouye K (2004) Purification and characterization of mouse CYP27B1 overproduced by an Escherichia coli system coexpressing molecular chaperonins GroEL ⁄ ES Biochem Biophys Res Commun 323, 505–511 Arase M, Waterman MR & Kagawa N (2006) Purification and characterization. .. scan and stopped-flow studies of carbonmonoxide binding to bovine adrenocortical cytochromeP-450scc Biochim Biophys Acta 911, 162–167 Lange R, Heiber-Langer I, Bonfils C, Fabre I, Negishi M & Balny C (1994) Activation volume and energetic properties of the binding of CO to hemoproteins Biophys J 66, 89–98 Tuckey RC & Kamin H (1983) Kinetics of O2 and CO Binding to adrenal cytochrome P-450scc Effect of. .. (2002) The human steroid hydroxylases CYP1B1 and CYP11B2 Biol Chem 383, 1537–1551 Schiffler B, Kiefer M, Wilken A, Hannemann F, Adolph HW & Bernhardt R (2001) The interaction of bovine adrenodoxin with CYP11A1 (cytochrome P450scc) and CYP11B1 (cytochrome P45 011b) acceleration of reduction and substrate conversion by site-directed mutagenesis of adrenodoxin J Biol Chem 276, 36225–36232 Bechtel S, Belkina N... cytochrome c-551 of Pseudomonas aeruginosa Biochim Biophys Acta 578, 392–400 Banci L, Bertini I, Eltis LD & Pierattelli R (1993) Spectroscopic characterization of a newly isolated cytochrome P450 from Rhodococcus rhodochrous Biophys J 65, 806–813 Belkina NV, Lisurek M, Ivanov AS & Bernhardt R (2001) Modelling of three-dimensional structures of cytochromes P450 11B1 and 11B2 J Inorg Biochem 87, 197–207 Gopalakrishnan... N & Bernhardt R (2002) The effect of amino-acid substitutions I112P, D147E and K152N in CYP11B2 on the catalytic activities of the enzyme Eur J Biochem 269, 1118–1127 Nishihara K, Kanemori M, Kitagawa M, Yanagi H & Yura T (1998) Chaperone coexpression plasmids: differential and synergistic roles of DnaK-DnaJ-GrpE and GroEL-GroES in assisting folding of an allergen of FEBS Journal 275 (2008) 799–810... Bernhardt R (2005) Phosphorylation of bovine adreno- 39 40 41 42 43 doxin by protein kinase CK2 affects the interaction with its redox partner cytochrome p450(scc) (CYP11A1) Biochemistry 44, 3821–3830 Zollner A, Pasquinelli MA, Bernhardt R & Beratan ¨ DN (2007) Protein phosphorylation and intermolecular electron transfer: a joint experimental and computational study of a hormone biosynthesis pathway... 2008 The Authors Journal compilation ª 2008 FEBS 809 Functional characterization of hCYP11B1 32 33 34 35 36 37 38 810 A Zollner et al ¨ Japanese cedar pollen, Cryj2 in Escherichia coli Appl Environ Microbiol 64, 1694–1699 Sagara Y, Takata Y, Miyata T, Hara T & Horiuchi T (1987) Cloning and sequence analysis of adrenodoxin reductase cDNA from bovine adrenal cortex J Biochem (Tokyo) 102, 1333–1336 Sagara . be explained by the broader substrate-binding spectrum of bCYP11B1 resulting from the combination of the functions of CYP11B1 and CYP11B2 within one enzyme. Comparison of the k cat values obtained. Purification and functional characterization of human 11b hydroxylase expressed in Escherichia coli Andy Zo ¨ llner 1 , Norio Kagawa 1,2 , Michael. volume and energetic properties of the binding of CO to hemoproteins. Biophys J 66, 89–98. 42 Tuckey RC & Kamin H (1983) Kinetics of O 2 and CO Binding to adrenal cytochrome P-450scc. Effect of